Determination of Additives in Polymers and Rubber

Determination of Additives in Polymers and Rubbers Roy Crompton Rapra Technology Shawbury, Shrewsbury, Shropshire, SY...

Author: TR Crompton

254 downloads 2580 Views 5MB Size Report

This content was uploaded by our users and we assume good faith they have the permission to share this book. If you own the copyright to this book and it is wrongfully on our website, we offer a simple DMCA procedure to remove your content from our site. Start by pressing the button below!
Report copyright / DMCA form

DOWNLOAD PDF

Determination of Additives in Polymers and Rubbers

Roy Crompton

Rapra Technology Shawbury, Shrewsbury, Shropshire, SY4 4NR, United Kingdom Telephone: +44 (0)1939 250383 Fax: +44 (0)1939 251118 http://www.rapra.net

HB Poly Additives.indb 1

21/3/07 11:21:25 am

First Published in 2007 by

Rapra Technology Shawbury, Shrewsbury, Shropshire, SY4 4NR, UK

©2007, Rapra Technology

All rights reserved. Except as permitted under current legislation no part of this publication may be photocopied, reproduced or distributed in any form or by any means or stored in a database or retrieval system, without the prior permission from the copyright holder. A catalogue record for this book is available from the British Library.

Every effort has been made to contact copyright holders of any material reproduced within the text and the authors and publishers apologise if any have been overlooked.

ISBN: 978-1-84735-036-7

Typeset, printed and bound by Rapra Technology Limited Cover printed by Livesey Ltd, Shropshire, UK

Contents

Contents Preface .................................................................................................................. xi 1

Direct Determination of Additives in Polymers and Rubbers .......................... 1 1.1

Infrared Spectroscopic Methods ............................................................. 2

1.2

Ultraviolet Spectroscopy ....................................................................... 11

1.3

Raman Spectroscopy ............................................................................ 16

1.4

Mass Spectrometry ............................................................................... 17

1.5

X-ray Photoelectron Spectroscopy (XPS) .............................................. 49

1.6

Thermal Methods of Analysis............................................................... 49 1.6.1

Differential Scanning Calorimetry ............................................ 49

1.6.2

Differential Thermal Analysis ................................................... 54

1.6.3

Thermogravimetric Analyses .................................................... 58

1.7

Vapour Phase Ultraviolet Spectroscopy ................................................ 58

1.8

X-Ray Fluorescence Analysis ................................................................ 58

1.9

Nuclear Magnetic Resonance Spectroscopy ......................................... 58

References ..................................................................................................... 61 2

Extraction Techniques for Additives in Polymers .......................................... 69 2.1

Introduction ......................................................................................... 69

2.2

Solvent Extraction ................................................................................ 71 2.2.1

Polyolefins ................................................................................ 71

2.2.2

Polystyrene ............................................................................... 75

2.2.3

Acrylic Polymers....................................................................... 76

2.2.4

PVC ......................................................................................... 76

iii

HB Poly Additives.indb iii

21/3/07 11:21:26 am


2.2.5

Rubbers .................................................................................... 82

2.2.6

Polyacrylamide ......................................................................... 82

2.2.7

Polyurethane ............................................................................ 82

2.2.8

Vinyl Chloride, Butadiene, Acrylonitrile, Styrene, 2 Ethylhexyl Acrylate Copolymers ........................................... 82

2.2.9

Other Polymers ........................................................................ 83

2.3

Fractional Precipitation ........................................................................ 83

2.4

Fractional Extraction............................................................................ 84

2.5

Separation by Diffusion Methods ......................................................... 85

2.6

Dialysis or Electrodialysis ..................................................................... 85

2.7

Vacuum Thermal Displacement Extraction Method ............................. 85 2.7.1

Effects of Polymer Milling on Extraction.................................. 87

2.8

Solvent Extraction – Infrared Spectrometry .......................................... 90

2.9

Solvent Extraction – Ultraviolet Spectroscopy ...................................... 96 2.9.1

Ionol in Polyolefins ................................................................... 96

2.9.2

Santonox R In Polyolefins ........................................................ 96

2.9.3

Styrene Monomer ................................................................... 101

2.10 Solvent Extraction – Visible Spectroscopy .......................................... 104 2.10.1 Phenol Antioxidants ............................................................... 104 2.10.2 Amine Antioxidants ............................................................... 107 2.10.3 Tris Nonyl (Phenylated Phenyl) Phosphite .............................. 109 2.11 Solvent Extraction Spectrofluorimetry ................................................ 110 2.11.1 Perkin-Elmer LS-2B Microfilter Fluorimeter ........................... 110 2.11.2 Antioxidants........................................................................... 111 2.12 Solvent Extraction – Mass Spectrometry ............................................ 114 2.12.1 Ultraviolet Absorbers ............................................................. 115 2.13 Solvent Extracts – Electrochemical Methods ...................................... 116 2.13.1 Acrylamide ............................................................................. 116

iv

HB Poly Additives.indb iv

21/3/07 11:21:27 am

Contents

2.13.2 Antioxidants........................................................................... 117 2.13.3 Organic Peroxides .................................................................. 118 2.13.4 Acrylonitrile ........................................................................... 124 2.13.5 Determination of Styrene........................................................ 126 2.13.6 Determination of Acrylonitrile ............................................... 126 2.13.7 Organometallic Stabilisers ...................................................... 126 2.14 Chronopotentiometry ......................................................................... 127 2.15 Anodic Voltammetry .......................................................................... 129 2.16 Solvent Extraction – Nuclear Magnetic Resonance Spectroscopy (NMR) .......................................................................... 130 References ................................................................................................... 131 3

Liquid Chromatography ............................................................................. 139 3.1

3.2

3.3

Introduction ....................................................................................... 139 3.1.1

The Isocratic System ............................................................... 145

3.1.2

Basic Gradient System ............................................................ 146

3.1.3

Advanced Gradient System ..................................................... 146

3.1.4

The Inert System..................................................................... 146

Chromatographic Detectors ............................................................... 147 3.2.1

Post-Column Derivatisation - Fluorescence Detectors............. 147

3.2.2

Diode Array Detectors............................................................ 150

3.2.3

Electrochemical Detectors ...................................................... 150

Antioxidants....................................................................................... 151 3.3.1

Instrumentation ...................................................................... 152

3.3.2

Applications ........................................................................... 153

3.4

Oligomers........................................................................................... 154

3.5

Acrylic Acid Monomer ....................................................................... 156

3.6

Acrylamide Monomer ........................................................................ 157

3.7

Amines ............................................................................................... 157

v

HB Poly Additives.indb v

21/3/07 11:21:27 am


3.8

Plasticisers .......................................................................................... 157

3.9

Additive Mixtures .............................................................................. 158

3.10 High Performance Liquid Chromatography – Infrared Spectroscopy .. 158 3.11 Gel Permeation Chromatography ....................................................... 159 References ................................................................................................... 160 4

Gas Chromatography ................................................................................. 165 4.1

Antioxidants....................................................................................... 165 4.1.1

Secondary Antioxidants.......................................................... 172

4.2

Volatile Compounds ........................................................................... 175

4.3

Monomers .......................................................................................... 179

4.4

Oligomers........................................................................................... 183

4.5

Hindered Amine Light Stabilisers (HALS) .......................................... 184

4.6

Plasticisers .......................................................................................... 185

4.7

Organic Peroxides .............................................................................. 197

4.8

Rubber Antidegradants ...................................................................... 199

4.9

Miscellaneous Polymer Additives ....................................................... 199

4.10 Identification of Additives by a Combination of GC and Infrared Spectroscopy ......................................................................... 201 4.11 Identification of Additives by a Combination of GC and Mass Spectrometry ............................................................................. 213 4.12 Pyrolysis GC....................................................................................... 215 References ................................................................................................... 216 5

Thin-Layer Chromatography ...................................................................... 225 5.1

Experimental ...................................................................................... 228 5.1.1

Preparation of Thin-layer Plates for Analysis.......................... 228

5.1.2

Application of Polymer Extract to Plate ................................. 229

vi

HB Poly Additives.indb vi

21/3/07 11:21:27 am

Contents

5.2

5.1.3

Selection of Chromatographic Solvent .................................... 230

5.1.4

Detection of Separated Compounds on the Plate .................... 232

5.1.5

Evaluation of Developed Plates .............................................. 236

5.1.6

Spectroscopic Methods ........................................................... 236

5.1.7

Optical Densiometric Analysis................................................ 238

5.1.8

Methods Based on Spot Size ................................................... 239

Antioxidants....................................................................................... 243 5.2.1

Determination of Santonox R................................................. 243

5.3

Ultraviolet Stabilisers.......................................................................... 248

5.4

Plasticisers .......................................................................................... 249

5.5

Organotin Stabilisers .......................................................................... 252

5.6

Epoxy and Other Heat Stabilisers....................................................... 254

5.7

Optical Whiteners .............................................................................. 256

5.8

Amine and Phenolic Antioxidants and Antidegradants, Guanidines and Accelerators in Rubber.............................................. 258

5.9

Miscellaneous Additives ..................................................................... 259

5.10 Combination of Thin-Layer Chromatography with Infrared Spectroscopy ......................................................................... 262 5.10.1 Premigration of Plates ............................................................ 263 5.10.2 Removal of Separated Compounds from the Plate.................. 265 5.10.3 Extraction of Pure Polymer Additives from Separated Adsorbent Bands .................................................................... 266 5.10.4 Preparation of Infrared Spectra Separated Additives............... 268 5.10.5 Preparation of UV Spectra of Separated Additives .................. 274 References ................................................................................................... 275 6

Paper Chromatography .............................................................................. 283 References ................................................................................................... 287

7

Supercritical Fluid Chromatography ........................................................... 289

vii

HB Poly Additives.indb vii

21/3/07 11:21:27 am


7.1

Antioxidants....................................................................................... 291

7.2

Oligomers........................................................................................... 298

7.3

Supercritical Fluid Chromatography-Mass Spectrometry (SFC-MS) ... 300

References ................................................................................................... 300 8

Headspace Analysis - Volatiles .................................................................... 305 8.1

Volatiles ............................................................................................. 305

8.2

Monomers .......................................................................................... 313

8.3

Oligomers........................................................................................... 314

8.4

Miscellaneous ..................................................................................... 314

References ................................................................................................... 314 9

Thermal Methods ....................................................................................... 317 9.1

Pyrolysis-Gas Chromatography-Mass Spectrometry ........................... 317

9.2

Evolved Gas Analysis ......................................................................... 320

References ................................................................................................... 330 10 Determination of Water .............................................................................. 333 References ................................................................................................... 335 11 Determination of Metals ............................................................................. 337 11.1 Destructive Techniques ....................................................................... 337 11.1.1 Atomic Absorption Spectrometry ........................................... 337 11.1.2 Graphite Furnace Atomic Absorption Spectrometry ............... 343 11.1.3 Atom Trapping Technique ...................................................... 345 11.1.4 Vapour Generation Atomic Absorption Spectrometry ............ 345 11.1.5 Zeeman Atomic Absorption Spectrometry.............................. 346 11.1.6 Inductively Coupled Plasma Atomic Emission Spectrometry .......................................................................... 350 11.1.7 Hybrid Inductively Coupled Plasma Techniques ..................... 353

viii

HB Poly Additives.indb viii

21/3/07 11:21:27 am

Contents

11.1.8 Inductively Coupled Plasma Optical Emission Spectrometry-Mass Spectrometry ........................................... 355 11.1.9 Pre-concentration Atomic Absorption Spectrometry Techniques ............................................................................. 357 11.1.10 Microprocessors .................................................................... 358 11.1.11 Autosamplers ........................................................................ 358 11.1.12 Applications: Atomic Absorption Spectrometric Determination of Metals ........................................................ 359 11.1.13 Visible and UV Spectroscopy ................................................. 366 11.1.14 Polarography and Voltammetry ............................................. 366 11.2 Non-destructive Methods ................................................................... 375 11.2.1 X-ray Fluorescence Spectrometry ........................................... 375 11.2.2 Neutron Activation Analysis .................................................. 381 11.2.3 Metal Stearate Stabilisers........................................................ 384 References ................................................................................................... 385 12 Non-metallic Elements ................................................................................ 391 12.1 Instrumentation .................................................................................. 392 12.1.1 Furnace Combustion Methods ............................................... 392 12.1.2 Oxygen Flask Combustion Methods ...................................... 396 12.2 Acid Digestions of Polymers ............................................................... 398 12.2.1 Chlorine ................................................................................. 398 12.2.2 Nitrogen ................................................................................. 401 12.2.3 Phosphorus ............................................................................ 401 12.2.4 Silica ...................................................................................... 401 12.3 X-ray Fluorescence Spectroscopy........................................................ 402 12.4 Antec 9000 Nitrogen/Sulfur Analyser ................................................. 403 References ................................................................................................... 403 Appendix 1 ....................................................................................................... 407

ix

HB Poly Additives.indb ix

21/3/07 11:21:27 am


Abbreviations ........................................................................................................... Index .......................................................................................................................

x

HB Poly Additives.indb x

21/3/07 11:21:28 am

Preface

Preface

This book is designed as a practical text for use in the laboratories of the plastic producer and user industries and by others such as universities and other institutions who are concerned with problems associated with additives and adventitious impurities in polymers, their breakdown mechanisms and their analysis. It is now about 30 years since the author wrote his first book on this subject and much has happened in the field since then. For example powerful new analytical tools have been made available to the chemist by a combination of various chromatographic techniques with methods of identifying separated additives and their degradation products by techniques based on infrared and mass spectrometry. In particular supercritical fluid chromatography combined with mass spectrometry has come to the fore. Combinations of polymer pyrolysis with gas chromatography with mass spectrometric identification of the pyrolysis products is throwing new light on what happens to antioxidants and other polymer additives during polymer processing and products life. Similarly evolved gas analysis and thermogravimetry and dynamic scanning calorimetry are proving very useful in studies of antioxidant loss during polymer processing and service life. The book is an up-to-date coverage of the present state of knowledge on the subject of polymer additive systems and as such should be extremely useful to workers in the field.

T Roy Crompton March 2007

xi

HB Poly Additives.indb xi

21/3/07 11:21:28 am


xii

HB Poly Additives.indb xii

21/3/07 11:21:28 am

Direct Determination of Additives in Polymers and Rubbers

1

Direct Determination of Additives in Polymers and Rubbers Author

In general, the direct determination of additives in plastics, as opposed to carrying out a preliminary extraction technique, such as is discussed in Chapter 2, is less time consuming and more reproducible. The direct determination of all the additives in such extracts is not always possible because of spectral interferences from other additives, low relative molecular weight (MW), mass matrix oligomers and the extracting solvent. Infrared (IR), and ultraviolet (UV) spectrometric techniques have been used successfully in some cases; in others where the extract is a complex mixture, prior chromatographic separation of the additives is necessary. Methods based on a preliminary extraction of additives from the polymer, then chromatographic separation before the analytical finish are obviously much more time consuming than methods based on direct analysis of the polymer. Much recent work on the development of direct methods has been carried out and is discussed next. Chemometrics is the art of extracting chemically relevant information out of data produced in chemical experiments with the use of mathematical and statistical methods [2]. The main issue is to structure the chemical problem in a form that can be expressed as a mathematical problem. Chemometrics have become an integral part of spectroscopy and other areas of chemistry. Calibration methods seek to express the dependent variable as a linear function of the independent variables. Linearisation of variables before calibration will make the calibration model less complicated. According to Beer’s law, the concentration and the film thickness are proportional to the absorbance. A transformation from transmittance to absorbance is therefore recommended. Varying film thicknesses might give a multiplicative effect to the absorption spectra. The effect of different thicknesses can be reduced either by scaling the data before calibration or by including the film thickness in the calibration as an extra variable. The unscrambler function multiplicative signal correction (MSC) is a normalisation function which calculates a multiplicative (B) and/or additive (A) factor for each sample to compensate for differences between the samples [1]: Mafter (i, j) = {Mbefore (i,j) - A(i)}/B(i)

(1.1)

1


21/3/07 11:21:28 am

Determination of Additives in Polymers and Rubbers where M(i,j) is the absorbance for sample i at wavenumber j before and after scaling. The scaling factors for each sample are calculated from the spectra in regions where spectra based on different concentrations of additives are believed to be approximately equal. Principal component analysis (PCA) is a statistical technique which, over the last decade, has become a regular tool for analysing chemical data [3-6]. If there is a relationship among any samples in a data set, the PCA will separate the samples into groups. Partial least squares (PLS) regression is often used for multivariate calibration [7-9]. PLS differs from other regression methods by using the dependent variable (concentrations) actively during the decomposition of the spectra. By balancing the information in the spectra and the related concentrations, the method reduces the impact of large, but irrelevant, variations in the spectra. For each variable, the calibration gives a linear regression equation of the form: [concs] = B0 + B1 A(λ1) + B2 A(λ2) + ... + Bn A(λn)

(1.2)

where A(λn) is the absorbance at wavenumber n, where n may represent a single wavenumber or an average. To discuss the prediction error, one must validate the calibration model [2]. There are two sorts of validation. One method is based on a new set of objects (external prediction). It requires a large and representative set of objects which have to be kept apart from the calibration for testing purposes only. The other validation method is based on the calibration data themselves (internal validation). In most cases, internal validation methods such as cross-validation and leverage correction [2] give sensible results with valuable information about the prediction ability. Cross-validation seeks to validate the calibration model with independent test data, but contrary to external validation it does not use data for testing only. The cross-validation is performed a number of times, each time with the use of only a few calibration samples as a test set. From the validation set it is possible to compare the prediction ability for the models, expressed by the estimated prediction mean square error.

1.1 Infrared Spectroscopic Methods Vigerust and co-workers [1] used multivariate calibration methods to establish a new method for measurement of three additives in low-density polyethylene (LDPE). The determination of the concentrations of silica, erucamide and butylated hydroxyl toluene (BHT) is based on infrared spectroscopy and a calibration model compared to traditional methods - this method is both time- and cost-effective and is more precise.

2


21/3/07 11:21:28 am

Direct Determination of Additives in Polymers and Rubbers A Perkin Elmer FTIR model 1710 was used to record spectra in the region 4000-400 cm-1. As shown in Figures 1.1-1.3, silica has intense broad bands in the region 650 to 450 and 1500 to 750 cm-1 with little interference or overlap from the polymer itself. BHT, which

Figure 1.1 Absorbance infrared spectra of pure polymer (A) and polymer 10,000 wt/ppm silica content (B) Reproduced with permission from B. Vigerust, K. Kolset, S. Nordenson, A. Henricsen and K. Kleveland, Applied Spectroscopy, 45, 173. ©1991, Society for Applied Spectroscopy, MD, USA [1],

Figure 1.2 Absorbance infrared spectra of pure polymer (A) and polymer with 9300 wt/ppm erucamide content (B) Reproduced with permission from B. Vigerust, K. Kolset, S. Nordenson, A. Henricsen and K. Kleveland, Applied Spectroscopy, 45, 173. ©1991, Society for Applied Spectroscopy, MD, USA [1],

3


21/3/07 11:21:28 am


Figure 1.3 Absorbance infrared spectra of pure polymer (A) and polymer with 3400 wt/ppm BHT content (B) Reproduced with permission from B. Vigerust, K. Kolset, S. Nordenson, A. Henricsen and K. Kleveland, Applied Spectroscopy, 45, 173. ©1991, Society for Applied Spectroscopy, MD, USA [1]

is a steric hindered phenol, has sharp bands around 3650 cm-1 and some small sharp bands in areas dominated by silica. Erucamide has broad bands in the region 1700 to 1600 and 3500 to 3100 cm-1. The infrared signals from BHT and erucamide are weaker, and both have overlapping regions with the polymer. The selective weighting functions used for the additives in the regression models were especially necessary in order to obtain good enough prediction results for BHT. But introduction of these functions will make the models more sensitive to noise when the important infrared signals are small. Thus, noise from the spectrophotometer in important infrared response areas would also be weighted and could disturb the prediction ability of the regression models. In contrast to silica, BHT and erucamide are reactive and can be partly consumed during extrusion and sample preparation. The results from the regression analysis are shown in Table 1.1. Best results were obtained for silica with a correlation coefficient (R2) of 0.99. For BHT the correlation coefficient was 0.84 and for erucamide 0.91. Polyethylene glycols (PEG) are used as antistatic agents in polyethylene (PE) resins. PEG is a difficult additive to analyse. It cannot be extracted either quantitatively or reproducibly. A simple, rapid and reliable method is required for PEG in PE. Kumar [10] has described a direct Fourier transform infrared (FTIR) spectrometric approach for successfully determining low concentrations (1000 kgf/mm2) were designed. These assemblies are mounted on the Harrick Seagull ATR attachment and measured by step-scan FTIR spectroscopy. The effect of static compression, air gaps, and refractive index changes were examined. Experimental and simulated results showed that the effect of air gaps between the sample and IRE and refractive index changes of the sample and IRE are negligible at values larger than a static torque of 40 cN-m and good signal-to-noise ratios and reproducible data can be obtained. Uniaxially and biaxially drawn polyethylene terephthalate (PET) films were measured by this method. Both bipolar and unipolar bands were observed in the dynamic in-phase ATR spectra, which can be associated with their micro-structural environmental changes. This technique shows promise in evaluating

10


21/3/07 11:21:30 am

Direct Determination of Additives in Polymers and Rubbers various polymer film materials, including biaxially oriented films, multilayer coated film surfaces, and molecular interactions between polymer-polymer and polymer-additives at the film surface. Various workers have reviewed the application of IR spectroscopy to the determination of additives [16, 17, 18]. Other recent applications of IR spectroscopy include the determination of slip agents in PE [19], ethyl acetate and ethanol in HDPE [20], stearic acid in polystyrene (PS) [21], talc, antimony trioxide and decabromophenylether flame retardants in polyvinyl chloride (PVC) [22-24], mould release agents [25] and binders in aged paint film [26].

1.2 Ultraviolet Spectroscopy This technique has found very limited applications in the direct analysis of additives in polymers. Soucek and Jelinkova [27] have also used this differential principle to determine in polypropylene (PP) two antioxidants (2,6-di-tert-butyl 4 methylphenol and 4-substituted 2,6-xylenol) which have virtually identical UV absorption spectra in the absence of alkali. The antioxidants can be distinguished in alkaline medium, where 4-substituted 2,6-xylenol forms phosphonate readily, thus allowing the utilisation of the bathochromic shift for its determination. The use of derivative spectroscopy reduces light scattering and matrix interferences when extracts from PP samples are measured. Lutzen and co-workers [28] describe an in-line monitoring, UV method for the determination of polymer additives such as thermal and UV stabilisers and antioxidants in polymers. Thermal UV spectroscopy has been used to identify and determine organic and inorganic pigments in polymers. Organic and inorganic pigments are used for coloration of polymers, polymer films and polymer coatings on metal containers. Vapour phase UV absorption spectrometry at 200 nm has been used to identify such pigments [29]. In this method powdered samples are directly vapourised in the heated graphite atomiser. Thermal UV profiles of organic pigments show absorption bands between 300 and 900 °C, while profiles of inorganic pigments are characterised by absorption bands at temperatures above 900 °C. Temperature, relative intensity, and width of the bands allow the identification of the pigments. The technique shows fast acquisition of thermal UV profiles (2-3 minutes for each run), good repeatability and wide thermal range (from 150 to 2300 °C). The method has been applied to a variety of polymers.

11


21/3/07 11:21:30 am

Determination of Additives in Polymers and Rubbers A practical example of the identification of pigments is given in Figure 1.7. A 1:1 mixture of organic pigment yellow (2-nitro-p-toluidine coupled with acetoacetanilide) and inorganic PY 34 (lead chromate) was vapourised using the conditions quoted in Figure 1.1(a). The thermal UV profile clearly shows two absorption bands at about 500 °C and 1250 °C. The first band is attributable to the vapours which originate from the decomposition and pyrolysis of the organic pigment, the second band corresponds to the decomposition and vapourisation of lead chromate at high temperature (mp = 844 °C). It is possible therefore to determine by a rapid run whether the pigment is a mixture or belongs to the organic or inorganic group. Blue and green organic pigments are commonly derivatives of copper phthalocyanine, which is characterised by remarkable resistance to thermal treatments. In fact these pigments show UV absorption at temperatures higher (about 800-900 °C) than those observed for yellow and red pigments Figure 1.7(b). Albarino [30] has stated that analysis of PE additives by means of UV spectroscopy is limited by excessive beam dispersion due to light scattering from the polymer crystalline regions. Additives at low concentrations (0.1%) require sample thicknesses which mean that analysis must be performed in the presence of a high level of scattering which may change unpredictably with wavelength. At lower levels of concentration and correspondingly greater sample thicknesses, unacceptable signal-to-noise ratios exist. Nevertheless, UV spectroscopy remains an attractive method for analysis of many additives. Principal advantages over IR analysis include greater sensitivity arising from higher extinction coefficients and a lack of interfering absorptions from the PE matrix. These advantages can be realised, however, only if background scattering from the polymer can be reduced. Albarino [30] demonstrated the feasibility of quantitative UV analysis of additives in PE at temperatures above the polymer melting point where the crystallites, which account for much of the scattering, are eliminated. Greater sample thickness and analytical sensitivity are possible compared to analysis of solid samples at room temperature. In this work, sample thickness was controlled by brass shims held between Suprasil grade silica windows (Heraeus Amersil, Inc.) by a faceplate bolted to the cell body. PE samples were prepared for analysis by calculating the weight required to fill the shim opening in the melt. Samples were inserted into the shim opening as pressed films cut to size; several layers were required for greater thicknesses. After gently tightening the faceplate, the cell was rapidly heated to 120 to 125 °C by supplying about 65 W to the heater. By proper tightening of the faceplate, the shim space was uniformly filled with PE, after which the cell was transferred to the sample compartment of a spectrometer. Upon warmup to the melt, an input power of 29 W maintained cell temperature within the limits given in Table 1.3 during scanning. Cell temperature was regulated only to the extent of maintaining the melt between 121 to 135 °C. A small temperature increase, given

12


21/3/07 11:21:31 am


Figure 1.7 Thermal UV profile of (a) 1:1 mixture of organic pigment yellow (2-nitro-ptoluidine coupled with acetoacetanilide and inorganic pigment PY 34 (lead chromate). (b) PB 15.4 copper phthalocyanine (beta form); PG 7 polychlorinated copper phthalocyanine (14-16 chlorine atoms) Thermal cycle Temperature, ºC Ramp times, s Hold temperature, ºC Time constant Source: Author’s own files

1 200 10 25

2 2000 8 10 0.5

3 2650 2 5

13


21/3/07 11:21:31 am


Table 1.3 Analysis of Irganox 1010 antioxidant in molten polyethylene Composition Irganox 1010 in PE (%)

Thickness Temperature Sample Baseline Antioxidant (cm) (°C) absorbance absorbance absorbance at 2800 Å at 2800 Å at 2800 Å

1

0.101

0.030

121-129

0.212

0.032

0.180

2

0.101

0.058

122-126

0.347

0.044

0.303

3

0.101

0.081

122-125

0.511

0.053

0.458

4

0.101

0.112

123

0.678

0.065

0.613

5

0.051

0.218

124-125

0.692

0.107

0.585

6

0.051

0.056

122-127

0.217

0.043

0.174

7

0.051

0.109

124-127

0.378

0.06

0.314

8

0.051

0.165

124

0.548

0.086

0.462

9

0.010

0.274

123-128

0.278

0.129

0.149

10

0.010

0.508

122-124

0.486

0.222

0.264

11

0.010

0.612

125-128

0.585

0.264

0.321

12

0.010

0.780

127-135

0.753

0.330

0.423

13

0

0.058

123-124

0.278

0.058

14

0

0.266

126-129

0.486

0.130

15

0

0.508

124-132

0.585

0.218

16

0

0.780

123-131

0.753

0.333

Reproduced with permission from R.V. Albarino, Applied Spectroscopy, 27, 1, 47. ©1973, Society for Applied Spectroscopy, MD, USA [30]

by the intervals of Table 1.3 was generally allowed. Spectra were found to be insensitive to temperature in the intervals 128 ± 4 °C to 145 ± 4 °C; a thermometer in contact with ‘woods metal’ (low melting point alloy) was used to indicate initial cell temperature and temperature upon completion of spectra. Possible temperature gradients across the PE melt were considered unimportant in view of the insensitivity of spectra to melt temperature. Micrometer measurements of thickness were made on the solidified PE samples. Errors due to polymer contraction on solidification were small, as the process of solidification generally results in a net volume change of the solid in the absence of constraints. As the polymer samples were not constrained in any dimension, contraction occurred along the length and width of the specimen as well as the thickness. That portion of the contraction resulting in a decrease in sample thickness was observed to be non-uniform across the face of the sample; micrometer measurements on this face were taken as true melt thickness.

14


21/3/07 11:21:31 am

Direct Determination of Additives in Polymers and Rubbers Shims designed to allow an outflow of excess molten polyethylene would facilitate thickness measurements as melt thickness would correspond to shim thickness. Albarino [30] used standards consisting of PE and Irganox 1010. These were made by milling at temperatures of about 127 °C. Samples containing 0.051 and 0.010% Irganox 1010 were made from a masterbatch containing 0.101% Irganox 1010. These standards and an unstabilised control were moulded into sheets 0.064 to 0.076 cm thick for use in the analysis. The effect of sample melting on scattering is illustrated in Figure 1.8. Figure 1.9 is the spectrum of a 0.045 cm PE specimen containing 0.101% Irganox 1010 at room

Figure 1.8 Direct UV spectra of 0.101% Irganox 1010 in polyethylene, A: 0.0045 cm, B: 0.058 cm, 122-126 ºC Reproduced from R.V. Albarino, Applied Spectroscopy [30]

15


21/3/07 11:21:31 am


Figure 1.9 Direct UV spectra of 0.051% Irganox 1010 in polyethylene. A: 0.056 cm, 122127 ºC, B: 0.109 cm, 124-127 ºC, C: 0.165, 124 ° C, D: 0.218 cm, 124-125 ºC Reproduced with permission from Albarino, Applied Spectroscopy [30]

temperature. Figure 1.9 was recorded at 122 to 126 °C with a 0.058 cm specimen. A very substantial decrease in scattering has resulted with little change in the antioxidant absorption at 2800 Å.

1.3 Raman Spectroscopy Fourier transfer near infrared Raman spectroscopy (400-10,000 cm-1) is useful for the examination of additives in polymer extracts [31].

16


21/3/07 11:21:32 am

Direct Determination of Additives in Polymers and Rubbers An example of the application of Raman spectroscopy is the identification of additives in fire retardant PP. When a sample of PP was examined by IR spectroscopy the strongest bands (9.8 and 14.9 µm) were due to a talc-type material and bands of medium intensity were assigned to PP and possibly antimony trioxide (13.4 µm). Additional weak bands in the 7.3-7.7 µm region were possibly due to decabromodiphenyl ether. In the Raman spectrum, however, the strongest bands (250 and 185 cm-1 shift) confirmed the presence of antimony trioxide and some bands of medium intensity confirmed the presence of decabromodiphenyl ether (doublet at 140, triplet at 220 cm-1 shift) and PP (800, 835, 1150, 1325, 1450 and 2900 cm-1 shift). The silicate bands that obscured the regions of the IR spectrum were not observed in the Raman spectrum. Although both of these spectroscopic methods have a wide use in their own right, this example demonstrates well the complementary value of the two methods, taking advantage of the fact that elements of high atomic number, e.g., antimony and bromine, have relatively more intense Raman spectra but the lighter elements show up clearly in the IR spectra. Other applications of Raman spectroscopy include monomers in polymethylmethacrylate [32] and additives in PVC [33].

1.4 Mass Spectrometry Mass spectrometry (MS) involves the study of ions in the vapour phase. This analytical method has a number of features and advantages that make it an extremely valuable tool for the identification and structural elucidation of organic molecules - including synthetic polymers: (a) the amount of sample needed is small (microgram level or less); (b) the molar mass of the material can be obtained directly by measuring the mass of the molecular (or quasimolecular) ion; (c) molecular structures can be elucidated by examining molar masses, ion fragmentation patterns, and atomic compositions determined by mass spectrometry; and (d) mixtures can be analysed by using ‘soft’ ionisation methods and hyphenated techniques (such as GC-MS, liquid chromatography-mass spectrometry (LC-MS), and MS/MS). Mass spectrometric methods are routinely used to characterise a wide variety of biopolymers, such as proteins, polysaccharides, and nucleic acids. Nevertheless, despite its advantages, MS has been under utilised in the past for studying synthetic polymer systems. It is fair to say that, until recently, polymer scientists have been rather unfamiliar with the advances made in the field of MS. However, MS in recent years has rapidly become an indispensable tool in polymer analysis, and modern MS today complements in many ways the structural data provided by NMR and IR methods. Contemporary MS of polymers is capable of changing the 17


21/3/07 11:21:32 am

Determination of Additives in Polymers and Rubbers protocols which have been established for years, for the molecular and structural analysis of macromolecules. Some of the most significant applications of modern MS to synthetic polymers are (a) chemical structure and end-group analysis, (b) direct measurement of molar mass and molar mass distribution, (c) copolymer composition and sequence distribution, and (d) detection and identification of impurities and additives in polymeric materials. In order to analyse any material by MS, the sample must first be vapourised (or desorbed) and ionised into the instrument’s vacuum system. Since polymers are generally nonvolatile, many mass spectral methods have involved degradation of the polymeric material before analysis of the more volatile fragments. Two traditional methods to examine polymers have been flash-pyrolysis GC-MS and direct pyrolysis in the ion source of the instrument. In recent years, however, there has been a marked tendency toward the use of direct MS techniques. While a continued effort to introduce MS as a major technique for the structural analysis of polymers has been made over the past three decades, MS analysts did not have a great impact upon the polymer community until the past five years or so. During this period outstanding progress has been made in the application of MS to some crucial problems involving the characterisation of synthetic polymers. Developments in two general areas have spurred this progress. Sector and quadrupole mass analysers, the traditional methods of separation of ions in MS, have recently been complemented by the development of powerful Fourier transform (FT-MS) and time-offlight (TOF-MS) instruments. The TOF analysers are particularly well-suited for detecting higher molar-mass species present in polymers. Parallel to this, new ionisation methods have been developed that are based on the direct desorption of ions from polymer surfaces. With the introduction of ‘desorption/ionisation’ techniques, it has become possible to eject large molecules into the gas phase directly from the sample surface, and thereby mass spectra of intact polymer molecules have been produced. The term ‘desorption/ionisation’ refers to a method in which the desorption/vapourisation and ionisation steps essentially occur simultaneously. Fortunately, the use of MS for polymer analysis took on a new dimension at the turn of the century. Up until the mid-1990s there was a steady - but not dramatic - increase in the number of journal publications on polymer mass spectrometry. Starting in 1995, however, there has been a marked increase in the number of polymer mass spectrometry reports in the literature. Also the number of symposia and conferences devoted to the subject has grown considerably in the last few years.

18


21/3/07 11:21:32 am

Direct Determination of Additives in Polymers and Rubbers The major reason for this increase has been the use of matrix-assisted laser desorption/ionisation-MS (MALDI-MS) for numerous polymer applications. MALDI is by no means the only mass spectral method that is useful for polymer analysis, but it has provided the impetus to get polymer people interested in what mass spectrometry can do. Hayes and Altenau [34] were the first to report the use of MS to directly characterise antioxidants and processing oil additives in synthetic rubbers. Since then, various MS techniques have been applied to the analysis of rubber and polymer additives either as extracts or on the sample surface by laser techniques as reviewed by Lattimer and Harris [35]. Lattimer reviewed the present situation regarding MS in polymer analysis [36]. Analysis of polymer extracts by MS has proved challenging. Electron impact mass spectra (EI-MS) are often difficult to interpret due to the high concentration of processing oils and the additives in the extract, and excessive fragmentation of the molecular ions. Desorption/ionisation techniques such as field desorption (FD) and fast atom bombardment (FAB) have been found to be the most effective means for analysing polymer and rubber extracts [37, 38]. FD-MS has proved to be a particularly useful technique, since molecular ion abundances are high with respect to fragmented ions [39]. Electrospray ionisation MS (ESI-MS and ESI-MS-MS) has also been used for the analysis of polymer additive mixtures [40]. Extraction and separation procedures are time consuming, rendering additive characterisation a slow and laborious process, there is the possibility that the extraction process may compromise the integrity of the additive mixture, leading to an inaccurate picture of polymer composition. Attempts at direct MS characterisation of additives in bulk polymer samples have centred on direct thermal adsorption of additives for the bulk polymer, followed by EI-MS, chemical ionisation (CI-MS) or field ionisation (FI-MS). However, this approach is linked to polymer additives that are stable or can provide meaningful fragment ions at elevated temperatures. Desorption/ionisation methods such as fast ion bombardment (FAB) [41], laser desorption [42, 43] and secondary ion MS (SIMS) have also been applied to the analysis of additives in bulk polymer samples. However, these single step techniques suffer to varying degrees from matrix interferences in the resulting mass spectra. Laser desorption/laser photoionisation time-of-flight MS (LPToFMS) is a technique that has great potential for the direct analysis of molecular species from complex host matrices. This two-step approach circumvents many of the problems, discussed previously, that have been encountered with other techniques. These various experimental techniques are discussed further next.

19


21/3/07 11:21:32 am


Earlier Experimental Techniques In 1986, Peltonen [44] applied MS to the identification of volatile breakdown products of heated PS. The early work includes that of Rudewicz and Munson [45] who vaporised Ionox 330 and Irganox 168 and UV 531 additives from PP in a heatable glass probe under chemical ionisation conditions using 1.4% ammonia in methane reagent gas. The dominant species in this mixture, NH4+, is a low energy reagent ion that reacts with the additives to give very simple spectra of (M + H)+ or (M + NH4)+ ions with little fragmentation. Lattimer and co-workers [46] have applied MS to the determination of organic additives (antioxidants and antiozonants) in rubber vulcanisates. Direct thermal desorption was used with three different ionisation methods (EI, CI, FI). The vulcanisates were also examined by direct FAB-MS as a means for surface desorption/ionisation. Rubber extracts were examined directly by the four ionisation methods. Of the vaporisation/ionisation methods, it appears that field ionisation is the most efficient for identifying typical organic additives in rubber vulcanisates. Other earlier applications include those of Bletsos and co-workers [47] who produced time-of-flight ion MS of additives in polydimethylsiloxane and polytetrafluoroethylene, MS of organic additives in carbon black filled styrene-butadiene rubber [48] and oxidative ageing of antioxidants present on polymer surfaces [49, 36]. In principle, the most straightforward way to identify organic additives in a compounded polymer is to heat the material to thermally desorb the volatile components. The evolved chemicals may then be directed into a MS ion source for analysis. EI is the traditional method of ionisation in MS, and in the early years of MS, it was the only method that was readily available. Several studies from the 1970s and the early 1980s describe heating rubber compounds (or their extracts) in vacuo to vapourise the components into an EI ion source [34, 50-52]. These studies were in general hampered by (a) extensive EI fragmentation and (b) intense signals from the processing oil that is contained in most rubber recipes. Later systemic studies by Lattimer described the use of ‘soft’ ionisation methods (CI and FI) in the direct analysis of model vulcanisates [46] as well as uncured compounds [48]. The resulting ‘survey’ spectra were much simpler in nature – and thus easier to interpret – than those obtained via EI-MS. All of these studies, which used single-stage MS methods, describe the identification of various organic ingredients in the elastomers [50-53]. In later work, tandem mass spectrometry (MS-MS) was shown to be effective for increasing the specificity and sensitivity of detection and identification of additives in direct rubber compound [54, 55]. Pyrolysis field ionisation (Py-FI-MS) was shown to be a good technique for analysis of both organic additives and rubber components in the 20


21/3/07 11:21:32 am

Direct Determination of Additives in Polymers and Rubbers same experiment [56]. Results of literature reports through 1989 were summarised in a review article on rubber-compound analysis [57]. Relatively few descriptions of direct mass spectral analysis of plastics compounds have appeared in the literature. In a rather early report, additives in PP compounds were thermally desorbed into a heated reservoir inlet for mass spectral analysis [58]. It was found that numerous stabilisers could be identified via 80 eV EI-MS. Thermal, desorption of additives via direct probe introduction of PP compounds was described in a later report [59]. A more recent paper considered the mass spectral analysis of both rubber and plastic compounds. This report was an overview, without much detail. Analysis of additives in PP compounds via direct thermal desorption CI-MS has also been described [45]. Ammonia as a reagent gas was found to yield very simple CI mass spectra. Finally, a recent report analysed a number of additives (antioxidants and light stabilisers) in PP compounds. Three ionisation methods (El, CI, FI) were used, and supplemental MS-MS and atomic composition (AC-MS) results were used for chemical structure elucidation/confirmation of various ingredients.

Soft Ionisation, Tandem (MS-MS) and High Resolution (Atomic CompositionMS) Mass Spectrometry Lattimer [60] has recently reported a method of the mass spectral identification of components (particularly residual volatile chemicals, organic additives, and degradation products) in a number of commercial elastomer compounds of unknown composition. Programmed direct probe heating of the compounded elastomer was used with three different methods of ionisation: 70 eV EI-MS, isobutane CI-MS, and FI-MS. It should be understood that both thermal desorption and pyrolysis are taking place. That is, residual volatile chemicals and most organic additives are thermally desorbed at lower temperatures (less than ~250 °C), while polymeric components are thermally decomposed (pyrolysed) at higher temperatures (greater than ~250 °C). In some cases, tandem mass spectrometry (MS-MS) and or high resolution mass analysis (for atomic compositions, AC-MS) was carried out to improve the specificity of the analysis. Lattimer gives examples to illustrate the various modern MS approaches that may be used in practical problem-solving applications. The first step in the analysis of an unknown elastomer compound is to obtain and examine low-resolution ‘survey’ MS from the material. These spectra cover a wide mass range (typically ~ 50-1000 Da) and give an ‘overview’ of the sample composition. The most useful ionisation method for this is generally field ionisation, since the simplest possible spectrum is obtained. In the FI spectrum molecular ions are dominant, which facilitates the characterisation of the complex organic additive mixtures that are present in typical elastomer compounds.

21


21/3/07 11:21:33 am

Determination of Additives in Polymers and Rubbers The first example is a competitive elastomeric (ethylene-propylene-diene terpolymer; EPDM) bearing used in an aerospace application. The rubber was examined by heating in the direct probe over the range 20-400 °C. Figure 1.10 is a composite FI-MS survey scan covering the sample heating range ~70-230 °C; the probe heating rate was 18 °C/min. For comparison, a composite 70 eV EI-MS covering the same heating range is shown in Figure 1.11. The EI spectrum is considerably more complex due to the large number of fragment ions from processing oil and other ingredients. After survey scans are acquired, the next step was to identify the molecular ions. In some cases the identities may be obvious from the molecular weights alone. In most cases involving unknowns, however, the supplemental techniques of MS-MS and/or AC-MS are used. The basic concept of MS-MS is to put two mass analysers (MS-1 and MS-2) in tandem. After passing through MS-1, the ions traverse a collision chamber, where a low-pressure gas induces decomposition (fragmentation). In the Finnigan MAT 95Q, MS-1 is a double-focussing (BE) mass analyser, and the collision chamber is contained in an octapole field to enhance transmission. The secondary fragment (or product) ions are then mass-separated in MS-2 and subsequently detected. In the Finnigan MAT 95Q, MS-2 is a quadrupole mass filter. The most common MS-MS experiment is the product ion scan. In this a precursor (or parent) ion of interest is selected (focussed) in MS-1 and decomposed in the collision cell; the resulting product (or fragment) ions are then massseparated in MS-2 and detected. Figure 1.12 is a typical product ion scan (EI-MS-MS) for a volatile component from the rubber bearing. The precursor ion is M+ 136, obtained by El ionisation. (Either El or CI is normally used for MS-MS experiments, since the ion current is more stable and intense compared to that obtained by FI-MS). The spectrum in Figure 1.12 closely resembles that of cumyl alcohol as found in libraries of standard El spectra. The principal product ions are m/z 121 (M - CH3)+ and m/z 43 (C2H3O+). To determine unambiguously the molecular formula (or atomic composition) of an ion of interest, the supplemental technique of ‘high resolution’ MS (AC-MS) was used. In this, masses of ions were measured accurately to three or four decimal places with increased resolution of the instrument. This is accomplished by ‘matching’ known reference peaks (usually CxFy+ ions from perfluorokerosene) with the unknown peaks in the sample. In the Finnigan MAT 95Q, this operation is facilitated with computerised peak matching algorithms. Masses measured with a few parts per million (ppm) accuracy can provide an unequivocal identification of the atomic composition. This is due to the fact that the atomic weights of the nuclides are not exact whole numbers on the 12C mass scale (e.g., 1H = 1.007825, 16 O = 15.99491, 14N = 14.00307, 32S = 31.97207). Accurate mass measurements are most often carried out with El or CI ionisation, since the signal-to-noise ratio, stabilities of ions, and availability of reference peaks are better in these modes (as compared to FI).

22


21/3/07 11:21:33 am


Figure 1.10 FI-MS survey scan of EPDM bearing (70-230 °C) Reproduced with permission from Lattimer and co-workers, Rubber Chemistry and Technology [60]

Figure 1.11 EI-MS survey scan of EPDM bearing (70-230 °C) Reproduced with permission from Lattimer and co-workers, Rubber Chemistry and Technology [60]

23


21/3/07 11:21:33 am


Figure 1.12 Product ion scan (El-MS/MS) of M+ 136 from EPDM bearing Reproduced with permission from Lattimer and co-workers, Rubber Chemistry and Technology [60]

In the case of the M+ 136 ion from the rubber bearing, an El accurate mass measurement gave the value of m/z as 136.0892. Computer calculations showed that the best atomic composition match for this mass is C9H12O, which has a calculated m/z of 136.0888. The observed value differs from the calculated number by 3 ppm, which is an acceptable match. With the combination of the product ion scan (fragmentation pattern) obtained by MS-MS and the formula obtained by accurate mass measurement, the MW 136 component can confidently be assigned to the cumyl alcohol molecule. In summary, the organic additives in this competitive EPDM bearing were found to be a light aliphatic processing oil, polytrimethyldihydroquinoline and Irganox 1076 antioxidants, and fatty acids/esters. The curing agent was cumyl peroxide. One additional peak of interest in the FI MS is m/z 108, which was found by EI-MS-MS to be residual methylnorbornene from the EPDM polymer. Similarly application of this methodology to a rubber V-shaped belt showed it contained paraffin wax, a light unsaturated oil (wax, antiozonant, disphenylamine/acetone resin antioxidant, fatty acids, rosin acids and N-t-butyl 2-benzothiazole (sulfenamide accelerator).

24


21/3/07 11:21:33 am


Sulfenamide Accelerators Other applications of the tandem MS-MS technique include, determination in rubbers of general additives [55, 60] and the determination in polymers of antioxidants [61] and acrylate, methylmethacrylate and butyl acrylate monomers in acrylic thermoplastics [62].

Laser Mass Desorption/Electron Ionisation MS Johlman and co-workers [63] compared laser desorption/ionisation FT-MS (LD-FTMS) with FAB spectra of the same materials in the analysis of non volatile polymer additives. Both a pulsed carbon dioxide laser and a neodymium-YAG laser with outputs of 10.6 and 1.064 µm, respectively, were used to obtain LD-FT-MS spectra of all samples. Three sterically hindered phenols and other additives containing a variety of functionalities including thioester, phosphite, phosphonite, and hindered amine groups were examined. In general, FAB spectra show undesirably large amounts of fragmentation, while molecular ion species dominate LD-FT-MS spectra. It is concluded that LD-FT-MS spectra are superior to FAB spectra for analysis of these common polymer additives. This is illustrated in Figure 1.13. In the FAB spectrum of dilaurylthiopropionate in Figure 1.13 only a small peak resulting from the potassium attached molecular species appears. Fragmentation is substantial and corresponds to cleavage of the ester links with an m/z of 329, which further fragments to yield prominent ions with m/z of 133, 144 and 161. Laser desorption spectra acquired by carbon dioxide (Figure 1.13b) and Nd:YAG (Figure 1.13c) laser absorption, contrast substantially with the FAB spectra. Abundant molecular ion species are observed in the laser absorption spectra. These ions are a combination of (M + H)+, (M + Na)+, and (M + K)+, depending upon the relative abundance of alkali metal salts present in the sample. Present in the CO2 spectra of both dilaurylthiodipropionate (DLTDP) in Figure 1.13b are fragment ions with m/z 329 and m/z 413 that correspond to ester cleavages, as observed in the FAB spectra. In contrast, Nd:YAG spectra of DLTDP in Figure 1.13c each primarily contain a strong ion signal corresponding to cation attachment to the intact molecules. In addition to dialkylthiopropionate secondary antioxidants a similar situation was shown to apply in the case of higher MW compounds with thioester functionalities, e.g., Seenox 4125 with a MW of 1156, also hindered phenols, alkyl phosphites are polyhindered amines and Irganox 1050. Waddell and co-workers [64] applied this technique to Neoprene rubber compound surfaces. The LD-MS of the sulfur-vulcanised natural rubber (NR) Compounds #1 and

25


21/3/07 11:21:34 am


Figure 1.13 Comparison of positive ion spectra of DLTDP acquired by (a) FAB, (b) CO2 LD-FTMS and (c) Nd:YAG LD-FTMS Reproduced from Johlman and co-workers, American Chemical Society [63]

26


21/3/07 11:21:34 am

Direct Determination of Additives in Polymers and Rubbers #2 obtained using the LAMMA 1000 at high laser power, essentially show a continuous series of peaks up to approximately a mass-to-charge ratio (m/z) of 250, and were relatively uninformative. Using reduced laser power and focusing on a fresh surface area, the spectrum shown in Figure 1.14 was obtained for Compound #2. Similar mass spectra were also recorded for the unfilled Compound #1 and the silica-filled Compound #3. Thus the peaks observed in Figure 1.14 are thought to result from extensive fragmentation of the NR backbone since the m/z 68 peak thought indicative of the isopropenyl ion (C6H8+) is present, but no peaks greater than approximately m/z 200 are present. The laser mass spectrum differs from that obtained using pyrolysis methods which show isoprene oligomers up to m/z 900. The following compounds were identified in a range of Neoprene samples (Table 1.4), NSA carbon black filler, Sundex 8125 processing oils, Wingstay 300 antiozonant, Wingstay 100 antioxidant and sulfur curing agent.

Figure 1.14 Laser desorption mass spectrum of Compound #2, the carbon-black filled vulcanised rubber compound, obtained using the LAMMA 1000 spectrometer Reproduced with permission from Waddell and co-workers, Rubber Chemistry and Technology [64]

27


21/3/07 11:21:34 am


Table 1.4 Rubber compound formulas (parts per hundred rubber, phr) Compound

1

2

3

4

5

6

7

8

100

100

100

100

100

100

100

100

ISAF carbon black

0

50

0

50

0

50

HI-SUL 233 silica

0

0

50

0

0

0

50

0

Sulfur

2

2

2

2

2

2

2

2

Zinc oxide

3

3

3

3

3

3

3

3

0.5

0.5

0.5

0.5

0.5

0.5

0.5

0.5

Sundex 8125 processing oil

0

0

0

4

0

0

0

0

Wingstay 300 antiozonant

0

0

0

0

4

4

4

2

Wingstay 100 antioxidant

0

0

0

0

0

0

0

2

Natural rubber

Mercaptobenzothiazole

ISAF Reprinted with permission from W. Waddell, K. Benzing, L. Evans, S. Mowdood, D. Weil, J. McMahon, R. Cody, Jr., and J. Kinsinger, Rubber Chemistry and Technology, 64, 4, 622. ©1991, Rubber Division, ACS [64]

Regarding the processing oil, GPC of an authentic sample of Sundex 8125 aromatic processing oil and of an extract of cured Neoprene rubber containing this oil (Figure 1.15) shows that the processing oil is a mixture having components with a broad distribution of MW. Using a hydrocarbon as the GPC standard, the average MW of the processing oil is assigned a value of 200, however, using PS as a standard, affords a MW of 340. The MW reported by the manufacturer is 395. The direct surface analysis of Compound #4 by LDMS (Figure 1.16) gives a spectrum having a series of mass peaks with values ranging from m/z of about 200 to 360, centred around an m/z value of approximately 260. These peaks are thought to be due to the molecular ions (M+) of the various components comprising the aromatic processing oil created by loss of an electron from the aromatic ring. The discrepancy in MW distribution of the oil from that reported might be due to the relative diffusion characteristics of the lower MW and more volatile components in the oil which can be expected to result in their higher rubber surface concentrations as determined by direct analysis of the compound by LD-MS. The ATR-IR spectrum of rubber Compound #6 that contains the antiozonant has one clearly visible additional peak at 1470 cm-1, Figure 1.17 (arrow). The IR difference spectrum (Compound #6 - Compound #2), in Figure 1.18 reveals approximately six peaks (checkmarks) thought to be characteristic of the added antiozonant that might be used for its identification in a cured rubber compound since these peaks are present in the IR of the antiozonant, Figure 1.18b (checkmarks).

28


21/3/07 11:21:34 am


Figure 1.15 Gel permeation chromatographs of the aromatic processing oil (solid line) and the rubber extract (dotted line) Reproduced with permission from Waddell and co-workers, Rubber Chemistry and Technology [64]

Figure 1.16 LAMMA 1000 spectrum of Compound #4, the carbon-black filled, vulcanised rubber compound containing the aromatic processing oil, Sundex 8125 Reproduced with permission from Waddell and co-workers, Rubber Chemistry and Technology [64]

29


21/3/07 11:21:35 am


Figure 1.17 Attenuated total reflectance Fourier transform infrared spectrum of Compound #6, the carbon-black filled, vulcanised rubber compound containing the antiozonant. The arrow highlights the only peak visibly different from the ATR-IR spectrum of Compound #2, which does not contain the antiozonant Reproduced with permission from Waddell and co-workers, Rubber Chemistry and Technology [64]

The LAMMA 1000 LD-MS of Compound #6 has five new peaks present at m/z values of 268, 253, 211, 183, and 168, Figure 1.19 (checkmarks). Peaks thought to be representative of polymer backbone fragmentation are present at m/z values less than about m/z 120, including the m/z 68 peak, thought to be due to the isopropenyl ion. These new peaks are thought to result specifically from laser desorption and ionisation of the aromatic antiozonant present on the rubber surface. The LD-MS of Compound #8, which contains the Wingstay 300 antiozonant and an aromatic antioxidant, has characteristic peaks at m/z 268, 211, and 183 representative of the antiozonant and new peaks present at m/z 352, 288, 274, and 260. These latter three peaks are thought to represent the three molecular ions of the components of the antioxidant mixture in Goodyear’s Wingstay 100, an aromatic amine antioxidant. LD-MS has proven a uniquely useful technique for the direct characterisation of rubbercompound surface species. Mass spectra were obtained for intact molecular ions (M+) of organic chemical rubber additives such as the aromatic processing oil, and the aromatic antiozonant and antioxidants incorporated to protect the rubber. MW information from

30


21/3/07 11:21:35 am


Figure 1.18 A: Infrared difference spectrum obtained by computer subtraction of the ATR-IR spectra of Compound #2 from Compound #6, B: Infrared transmission spectrum of a thin film of the antiozonant on a NaCl plate. Checkmarks highlight those peaks characteristic of the para-phenylenediamine antiozonant Reproduced with permission from Waddell and co-workers, Rubber Chemistry and Technology [64]

the molecular ions and structural information from the fragmentation ions could be obtained without interference from the fragmentation peaks of the rubber backbone. Laser and thermal desorption mass spectral techniques provided complementary structural information, and when coupled with current analytical methods to characterise rubber compounds, can provide the necessary information to positively identify various organic species present on the surfaces of vulcanised rubber.

31


21/3/07 11:21:35 am


Figure 1.19 LAMMA 1000 spectrum of Compound #6, the carbon-black filled, vulcanised rubber compound containing the antiozonant. Checkmarks highlight those peaks characteristic of the para-phenylenediamine antiozonant Reproduced with permission from Waddell and co-workers, Rubber Chemistry and Technology [64]

Wright and co-workers [61] have used a two-step laser desorption/laser photoionisation time-of-flight MS (L2MS) for selective in situ detection of polymer additives in PP and polyoxymethylene. A pulsed CO2 laser was used to desorb the additives as neutral species into the gas phase, where they were post-ionised using a second UV laser operating at either 266 or 193 nm. For all the antioxidants studied, the 266 nm photoionisation mass spectra are dominated by the molecular ion peak; very little fragmentation is observed. In contrast, at 193 nm, the molecular ion peak is usually absent from the photoionisation mass spectra. Similar behaviour is exhibited by the UV stabilisers (Tinuvin) in their photoionisation mass spectra. This wavelength-dependent fragmentation can be exploited for unambiguous identification of many polymer additives. For example, it is shown that the isomeric UV stabilisers Tinuvin 320, Tinuvin 343, and Tinuvin 329 can be differentiated on the basis of

32


21/3/07 11:21:35 am

Direct Determination of Additives in Polymers and Rubbers the extent or nature of the observed fragmentation in their photoionisation mass spectra. Several commercial polymer formulations containing these types of additives have also been analysed using this experimental approach: the samples were interrogated directly without any pretreatment or extraction. It is shown that UV laser post-ionisation enables selective detection of the additives in preference to the polymer, providing unambiguous in situ identification. The potential of this technique for surface analysis and depth profiling is also discussed. There are many advantages to be gained in being able to chemically speciate additives directly from the polymer matrix as opposed to methods involving a preliminary solvent extraction of the additives from the polymer prior to MS (see Chapter 3) [43, 61, 65-73]. Laser desorption/laser photoionisation time-of-flight mass spectrometry is a technique that has great potential for the direct analysis of molecular species from complex host matrices. This two-step approach circumvents many of the problems that have been encountered with other techniques. In this method, a pulsed CO2 laser is used to desorb the analyte into the gas phase as a neutral species, directly from the sample of interest. A second pulse from a UV laser is then used to post-ionise these gas phase neutral species, generally using a resonance-enhanced multiphoton ionisation (REMPI) scheme. The benefits of this twostep approach lie in the spatial and temporal separation of the desorption and ionisation events, thereby enabling the independent optimisation of each process. This provides a number of advantages for the in situ analysis of bulk polymer samples: (i) desorption of neutral target molecules from the host polymer matrix with minimal decomposition, (ii) soft ionisation of the desorbed neutral species, resulting in readily interpretable mass spectra, (iii) selective ionisation of polymer additives which have a significant one-photon absorption cross section at the chosen ionisation wavelength, and (iv) highly sensitive detection of many polymer additive species. The two different ionisation laser wavelengths result in markedly different mass spectra. These mass spectral differences are a valuable aid in the unambiguous identification of the additives. Wright and co-workers [61] also reported that the spectra obtained show not only that it is possible to directly detect these additives in the polymer formulations, but also that chemical changes undergone by antioxidants, due to either processing or ageing, can also be observed. The additives included in this study are shown in Table 1.5. Antioxidants. The mass spectra obtained for Irganox 1330, Irgafos 168, and Santowhite using 266 and 193 nm photoionisation are shown in Figure 1.20. In the case of Irganox 1330, the spectra at both these wavelengths are dominated by the molecular ion peak at m/z = 774. However, it is evident that 266 nm photoionisation results in the production of fragment ions different than those observed at 193 nm. At 266 nm, a fragment is observed

33


21/3/07 11:21:36 am


Table 1.5 Nomenclature and structure of polymer additives Trivial name

Chemical name

Structure

Molecular weight

Irganox 1330 1 ,3,5,-tris(3,5-di-tert-4butyl-4-hydroxybenzyl)2,4,6,trimethylbenzene

774

Irganox 1076 Octadecyl-3-(3,5-di-tert-butyl4-hydroxyphenyl) propionate

530

Irgafos 168

Tris(2,4-di-tert-butylphenyl) phosphite

646

Santowhite

4,4′-Butylidene bis-6-(4-methyl2-tert-butylphenol)

382

Tinuvin P

2-(2′-Hydroxy-5′-methylphenyl)2H-benzotriazole

225

Tinuvin 326

2-(2′-Hydroxy-3′-methyl5′-tert-butylphenyl)-2H-5chlorobenzotriazole

315

Tinuvin 327

2-(2′-Hydroxy-3′,5′-ditert-butylphenyl)-2H-5chlorobenzotriazole

357

Tinuvin 320

2-(2′-Hydroxy-3′,5′-di-tertbutylphenyl)-2H-benzotriazole

323

34


21/3/07 11:21:36 am


Table 1.5 Cont’d… Trivial name

Chemical name

Structure

Molecular weight

Tinuvin 343

2-(2′-Hydroxy-3′-tert-butyl-5′(1-methyl)-propylphenyl)-2Hbenzotriazole

323

Tinuvin 329

2-(2′-Hydroxy-5′-(1,1,3,3di-methyl)-butylphenyl)-2Hbenzotriazole

323

Reproduced with permission from S.J. Wright, M.J. Dale, P.R.R. Langridge-Smith, Q. Zhan and R. Zenobi, Analytical Chemistry, 68, 20, 3585. ©1996, ACS [61]

Figure 1.20 Photoionisation mass spectra for Irganox 1330 at (a) 266 and (b) 193 nm; Irgafos 168 at (c) 266 and (d) 193 nm; and Santowhite powder at (e) 266 and (f) 193 nm. The peak marked with an asterisk in spectrum (c) is due to an internal mass standard, 4-aminobenzoic acid (m/z = 137) Reproduced with permission from S.J. Wright, M.J. Dale, P.R.R. Langridge-Smith, Q. Zhan and R. Zenobi, Analytical Chemistry, 68, 20, 3585. ©1996, ACS [61]

35


21/3/07 11:21:36 am

Determination of Additives in Polymers and Rubbers at m/z = 556. This is thought to result from loss of a 3,5-di-tert-butyl-4-hydroxybenzyl side group, with a concomitant hydrogen rearrangement. A weaker fragment peak at m/z = 57, due to the tert-butyl ion, can also be seen. Photoionisation of Irganox 1330 at 193 nm produces fragments at m/z = 57, 219, and 569. The fragment at m/z = 569 corresponds to the loss of a 3,5-di-tert-butyl-4-phenol side group via direct cleavage rather than rearrangement. The peak at m/z = 219 is characteristic of positive ion mass spectra of dibutyl phenols and corresponds to the 3,5-di-tert-butyl4-hydroxybenzyl ion. The peak at m/z = 57 corresponds to the tert-butyl ion. Similar characteristics were observed in the mass spectra for Irganox 1076 (not shown). When 266 nm radiation is used, the molecular ion can be clearly identified at m/z = 530. At 193 nm, no molecular ion peak is seen. Instead, the base peak of the mass spectrum is at m/z = 515, corresponding to the loss of a methyl radical from the molecular ion. In summary, for all the antioxidants studied, the mass spectra obtained using photoionisation at 266 nm are dominated by molecular ion signals, with very little fragmentation. With the exception of Irganox 1330, photoionisation using 193 nm radiation generated little or no molecular ion signal. There are two possible explanations for this apparent wavelength dependence. Most organic molecules have ionisation potentials (IP) in the range 7-10 eV. The energy of 266 nm photons is ~4.66 eV, whereas a 193 nm photon has an associated energy of 6.42 eV. In both cases, therefore, absorption of two photons is required in order to achieve ionisation. However, absorption of the first 193 nm photon can result in different intermediate electronic states being accessed compared to excitation at 266 nm, such as Rydberg states. It is possible that this may lead to different ionisation pathways being promoted, resulting in differing mass spectra at 193 versus 266 nm. Alternatively, the difference may simply be due to the difference in excess energy deposited initially in the molecular ion. Assuming an ionisation potential of 8 eV, photoionisation at 266 nm will produce a molecular ion with up to 1.32 eV of excess energy. This is small compared to the excess energy of up to 4.84 eV possible following photoionisation at 193 nm. This larger excess energy may be sufficient to exceed the appearance potential for the production of the most facile fragment ions. Therefore, at 193 nm, ionisation would be accompanied by facile fragmentation. The data available do not permit identification of which of these two mechanisms may be responsible for the different fragmentation patterns observed.

UV Stabilisers Similarly 266 and 193 nm photoionisation mass spectral data showed that in general the photoionisation mass spectra of the Tinuvin UV stabilisers examined differ markedly at 266 and 193 nm. At 266 nm, the mass spectra are dominated by molecular ion signals, 36


21/3/07 11:21:37 am

Direct Determination of Additives in Polymers and Rubbers with very little associated fragmentation. This nicely illustrates the advantage of L2MS as a soft ionisation technique, enabling readily interpretable mass spectra to be generated. Photoionisation at 193 nm, however, results in mass spectra in which the base peaks are fragment ions. This difference in behaviour may be due either to the difference in excess energy deposited in the molecular ion or to excitation via different intermediate states, as discussed earlier. Clearly, any technique that can provide chemical analysis of target analytes at trace levels directly from their host matrix represents an attractive and rapid methodology. The feature of the technique that allows this to be achieved is the selectivity provided by the photoionisation process. Most organic molecules have ionisation potentials between 7 and 10 eV. To achieve ionisation, absorption of two or more UV photons is required. For efficient photoionisation, a molecule must have a significant absorption cross section at the wavelength of the ionising radiation used. Molecules that do not possess a suitable chromophore will not be efficiently ionised. Therefore, by careful choice of the ionisation laser wavelength, the target analyte of interest may be selectively detected in preference to other components present in the mixture, including the host matrix. Wright and co-workers [61] showed that polymer additives with an appreciable absorption in the UV region of the spectrum can be selectively ionised in preference to the non-UV-absorbing host polymer. They examined a sample of PP containing 0.15 wt% of Irganox 1330 and 0.05 wt% of Irgafos 168. The mass spectra obtained using 266 and 193 nm photoionisation following direct desorption from the PP matrix are shown in Figure 1.21. To obtain an appreciable signal from this sample, it was necessary to increase the desorption laser power density 4-fold, to ~38 MW/cm2, compared to the value used to desorb the pure polymer additives. The need for increased laser desorption power densities is due to the PP matrix having a very low absorbance at the desorption laser wavelength of 10.6 nm. In the spectrum obtained at 266 nm, the molecular ions for both Irganox 1330 (m/z = 774) and Irgafos 168 (m/z = 646) are present. For ionisation at 193 nm, the molecular ion signals are very much weaker. However, a strong characteristic fragment signal at m/z = 441, anticipated from the 193 nm photoionisation mass spectrum of pure Irgafos 168 (see Figure 1.20d), can be seen. This peak is due to the loss of a 2,4-di-tert-butylphenyl—O group from the molecular ion, as is seen in the corresponding spectrum for the pure compound. These spectra demonstrate that, by use of two readily available ionisation wavelengths, and with reference to the corresponding spectra for the pure additives, it is possible to unambiguously determine the presence of Irganox 1330 and Irgafos 168 directly from the host PP matrix. An apparently anomalous peak at m/z = 662 is observed in the 266 nm photoionisation MS (see Figure 1.21a) which is due to a phosphate antioxidant which is generally due to an oxidation product of the Irgaflox 168 phosphite secondary antioxidant, i.e., it is possible to determine not only the active phosphate level in the polymer but also the 37


21/3/07 11:21:37 am


Figure 1.21 In situ mass spectra of polypropylene (PP) sample containing Irganox 1330 (0.15 wt%) and Irgafos 168 (0.05 wt%). Photoionisation at (a) 266 and (b) 193 nm. [M1]+ and [M2]+ denote the molecular ions of Irganox 1330 and Irgafos 168, respectively Reproduced with permission from S.J. Wright, M.J. Dale, P.R.R. Langridge-Smith, Q. Zhan and R. Zenobi, Analytical Chemistry, 68, 20, 3585. ©1996, ACS [61]

inactive degraded phosphate level, i.e., it is possible not only to determine the presence of additive species directly from the polymer but also to monitor chemical changes caused by the polymerisation process or subsequent exposure to heat, light, and other conditions which initiate polymer degradation.

38


21/3/07 11:21:37 am

Direct Determination of Additives in Polymers and Rubbers In conclusion, Wright and co-workers [61] have shown that marked differences in the photofragmentation behaviour at 266 and 193 nm, allow unambiguous identification of these additives, even including differentiation between isomeric species. It has also proved possible to detect antioxidants and UV stabilisers in PP and polyoxymethylene polymers at concentrations consistent with commercial polymer formulations. In the case of the PP polymer formulation, it has been possible to detect an oxidation product of the antioxidant Irgafos 168 formed during either processing or natural ageing of the polymer. Such measurements could be extended to allow monitoring of additive degradation levels in aged polymer samples. This study has also demonstrated the potential of L2MS as a surface analytical technique. It has been shown that it is possible to detect species on the surfaces of polymers which are not present in the bulk of the sample. It should prove possible to extend this work using spatially resolved desorption to probe for additive migration and aggregation. Zhan and co-workers [65] have also reported on the application of two-step laser MS to the determination of surface antioxidants and the UV stablisers on PE and PET. They showed that laser melting-depth profiling could be achieved in polyoxymethylene, which enabled the determination of the special distribution of an antioxidant in an injection moulded test bar.

Laser Desorption EI-FT Ion Cyclotron MS (LD-EI-FT-ICR-MS) As industrial thermoplastics are melt processed, they undergo oxidation reactions leading to changes in molecular weight and colour. Phosphite antioxidants [66-69] are generally considered to be secondary antioxidants, and their function is to control polymer molecular weight and colour. Phosphite stabilisers are used in most common thermoplastics at levels from 250 ppm to 2%. Typical phosphite loadings are often less than 1000 ppm. Phosphite stabilisers react with hydroperoxides, peroxy radicals, alkoxy radicals, and olefinic and carbonyl moieties; in addition, phosphites form co-ordination complexes with metals, changing their potential activity [70, 71]. Since the additive level in the polymer affects its stability, the analysis of polymer additives immediately poses two basic analytical questions: first, how much of the additive gets into the polymer during compounding, and second, how much of the additive that was added remains as the original phosphite form? Conventional methods for isolation and detection of phosphite additives are generally unsatisfactory. Phosphite antioxidant tends to fragment extensively in mass spectrometric analysis. Xiang and co-workers [72] analysed Ultranox 626 diphosphite [bis(2,4-di-tert-butylphenyl) pentaerythritol diphosphite] and its corresponding diphosphate oxidation product, XR-2502, as well as the phosphate additive, WESTON 618 diphosphite (distearyl pentaerythritol diphosphate) by Nd:YAG laser desorption 1.064 nm) electron ionisation Fourier transform

39


21/3/07 11:21:37 am

Determination of Additives in Polymers and Rubbers ion cyclotron resonance mass spectrometry (LD/EI/FT/ICR/MS). For each of the isolated additives, the molecular ion (M+) was observed as the predominant species with virtually no fragmentation. Moreover, abundant molecular ions were detected for Ultranox 626 diphosphite in a mixed polymer of polyethylene terephthalate, PP, and acrylonitrilebutadiene-styrene (ABS) at additive concentrations as low as 0.1% by direct analysis of the polymer film when the probe was heated to about 200 °C prior to laser desorption. The elevated sample temperature appears to increase the free volume of the polymer, in turn facilitating release of laser desorbed/ionised additives. LD/EI/FT/ICR/MS thus offers a sensitive and accurate means for detecting nonvolatile phosphite additives at typical concentrations in solid polymers, without the need for any chemical pretreatment.

Mass Spectra of Pure Additives The LD/EI/FT/ICR mass spectrum of the Ultranox 626 disphosphate additive is shown in Figure 1.22a. The molecular ion (M+) at mass-to-charge m/z 604 is the principal ionic species, in contrast to the pseudomolecular (M+ K+) ion observed in highest abundance for other kinds of additives [43, 73] by LD/FT/ICR/MS (i.e., no electron ionisation following laser desorption). By optimising the laser power (to ~50 mJ in ~10 ns in this case), it is possible to generate abundant molecular ions with virtually no fragmentation - higher laser power induces significant fragmentation. Scheme 1.1 shows the two-stage oxidation of Ultranox 626 diphosphite to form the diphosphate compound, XR 2502. The LD/EI/FT/ICR mass spectrum in Figure 1.22b shows the predominant molecular ion signal (m/z 636) for the diphosphate, XR 2502, along with residual signals from incompletely oxidised phosphite precursor at m/z 620 (half-oxidised monophosphate) and m/z 604 (unoxidised phosphite).

Scheme 1.1

40


21/3/07 11:21:38 am


Figure 1.22 LD/EI/FT/ICR mass spectra of ULTRANOX 626 diphosphite in various solid polymers: (a) 0.1% in PP; (b) 0.25% in acrylonitrile-butadiene-styrene; (c) 10% in PET. Direct analysis of polymeric film or extraction was necessary to produce these spectra. Reproduced from Xiang and co-workers, American Chemical Society [72]

Figure 1.22c shows the LD/EI/FT/ICR mass spectrum of a second phosphite additive, WESTON 618 diphosphite. Although the molecular ion (m/z 732) is readily observed, its abundance is lower than for Ultranox 626 diphosphite or XF 2502, presumably because WESTON 618 diphosphite has saturated C18 hydrocarbon chains in place of aromatic rings and therefore fragments more easily. Similar behaviour has been reported for other phosphate additives [43]. Figure 1.22a shows the LD/EI/FT/ICR mass spectrum of Ultranox 626 diphosphite (~0.1% w/w) in PP. Although no fragment ions from the polymer itself are observed under these conditions (due to low laser power and high MW of the polymer), molecular ions from the additive in both diphosphite (m/z 604) and diphosphate (m/z 636) form are clearly detectable at ~ 0.05% each. The signal magnitude increases significantly when the probe is heated to about 200 °C prior to laser desorption. Heating evidently increases the free 41


21/3/07 11:21:38 am

Determination of Additives in Polymers and Rubbers volume of the polymer to facilitate laser desorption/ionisation of the additives. Abundant ions at m/z 191 and 205 correspond to the following fragments:

The fragment at m/z 205 is an obvious phosphite bond cleavage product, the other fragment was confirmed to be C13H19O (191.143 nm) by accurate-mass measurement (191.145 nm) by internal calibration against ions of seven m/z ratios from perfluorotri-n-butylamine. LD/EI/FT/ICR mass spectra of the same additive present at higher concentrations in PET and ABS polymers are shown in Figure 1.22b and c. This time, there does not appear to be significant oxidation of the additive, since no signals from the phosphate or diphosphate oxidation products are observed. Laser desorption/Fourier transform ion cyclotron resonance MS has also been used to identify and determine the following types of polymer additives [74]: UV absorbents, e.g., Tinuvin [75], antioxidants, e.g., Irganox MD-1024, and amide wax antislip additives.

Potassium Ionisation of Desorbed Species (K+IDS) Potassium ionisation of desorbed species (K+IDS) with mass spectrometric detection is an extremely useful tool for the characterisation of high performance organic coatings. K+IDS uses a commercial rapid heating probe to desorb intact molecules which are then ionised by potassium cation attachment. Based upon the molecular ions, which appear as [M]K+, coatings components can be qualitatively and quantitatively analysed. In this work K+IDS was selected as a method of soft ionisation, (i.e., producing molecular ions) because of its simplicity, wide applicability, low cost and compatibility with the quadrupole mass spectrometer. Simonsick [76] reports the application of K+IDS to polymer additives (UV stabilisers and antioxidants), catalysts (organotin), reactive diluents (vernonia oil and aliphatic epoxides) and polyurethane precursors (polyesters and isocyanates). Tikuisis and co-workers [77] also discussed this technique. The technique is based upon rapid heating to desorb intact compounds [78]. Since ions are produced by potassium attachment and the internal energy transfer is low, primarily potassiated molecular ions with little or no fragmentation are observed [74, 79, 80]. Implementation of K+IDS required no modification of existing commercial equipment, no capital investment and is performed on quadrupole mass spectrometers [75]. 42


21/3/07 11:21:38 am

Direct Determination of Additives in Polymers and Rubbers This method was applied to the determination of organotin catalysts, e.g., dibutyl tin dilaurate, stablisers, e.g., Cyanox 1790, a hindered phenol antioxidant and Tinuvin 292, a hindered amine light stabiliser (HALS) and reactive diluents (such as aliphatic epoxides, epichlorohydrin, vernonial oil and so on) in acrylate resins. Figure 1.23 shows the K+IDS MS spectrum of Cyanox 1790 antioxidant. Figure 1.23 is the K+IDS mass spectrum of Cyanox-1790 a commercially available hindered phenolic antioxidant (American Cyanamid, Wayne, NJ). The molecular weight of this chemical is 699 Da; hence, under K+IDS conditions we would expect an ion at 738 Da, i.e., 699 + 39 (39 being the atomic weight of potassium). We see from the spectrum that this material is relatively pure. This paper demonstrates the utility of K+IDS for the characterisation of individual coating components. Molecular weight data and complementary isotope patterns permits a rapid (10 minutes) assignment of specific structures to the materials contained in coatings. A change in mass is usually involved in an organic reaction. The products differ from the reactants in molecular weight. Hence, K+IDS is an excellent probe for monitoring the success of derivatisation reactions and for elucidating the reactions which occur in model

Figure 1.23 K+IDS MS of Cyanox-1790, a commercial antioxidant Reproduced from W.J. Simonsick Jr., Progress in Organic Coatings, 1992, 20, 3-4, 411 [76]

43


21/3/07 11:21:38 am

Determination of Additives in Polymers and Rubbers crosslinking chemistries. Finally, using selective chemical degradation coupled with K+IDS analysis of the products, one is able to assemble the original network structure.

Fast Atom Bombardment (FAB) Sterically hindered phenols and other additives containing thioesters, phosphonites and hindered amine moieties were analysed by LD-FT-MS and FAB-MS [81]. The LD technique was preferred for analysis of polymer additives because of undesirable fragmentation from FAB techniques. Chang and co-workers applied FAB-MS to the measurement of HALS in polymers. The technique has also been applied to the measurement of oligomers up to decamer in polymers [82].

High Frequency Collision Induced Dissociation (CID) This technique has been used to analyse the effect of internal energy deposition on the collision induced dissociation (CID) fragmentation spectra of Irganox 1076. Four different ionisation techniques were compared. The variation in the relative yields of the different fragmentation ions was attributed to differences in the amount of internal energy transferred to the precursor ions during the ionisation process. A five component mixture of antioxidants and UV stabilisers has been analysed by high energy MS and high energy CID using a four sector instrument and a time-of-flight MS [83] and by electrospray ionisation with high energy CID [84].

Secondary Ion Mass Spectrometry (SIMS) The principles of this technique have been discussed by Shick and co-workers [85] and O’Toole and co-workers [86]. Time-of-flight SIMS with either gallium or indium primary beams has been evaluated as a method for measuring the homogeneity of distribution of a hindered amine antioxidant in low-density polyethylene. The parent ion for the oligomer at m/z 599 was so weak that it could not be used to map the distribution of the additive throughout its most commonly used concentration range (0.1 to 0.5% w/w) in polyethylene. Instead a mass fragment at m/z 58 was found to be sufficiently clear of interferences for use as a surrogate for the parent ion. As a result, imaging of the antioxidant distribution was possible to concentrations as low as 0.1% and a linear concentration calibration curve was obtained. The use of an indium primary beam improved the correlation of the antioxidant. Furthermore, indium reduced the contribution from the polyethylene background at m/z 58 in relation to the total counts acquired. 44


21/3/07 11:21:39 am

Direct Determination of Additives in Polymers and Rubbers Rudewicz and Munson [45] used this technique for the direct determination of additives in PP. The technique has also been used to determine oligomers in polyacrylates, PEG, siloxanes and polycarbonates [87], polyglycols [88] and adhesion promoters, primers and additives in the surface of PET film [89], volatile antioxidants in styrene-butadiene rubbers [34, 50], mercaptobenzothiazole sulfenamide accelerator in rubber vulcanisates [90] and divinyl benzene in styrene-divinyl benzene copolymer [91].

Maldi Mass Spectrometry Hanton [92] applied MALDI spectroscopy and electrospray ionisation MS to the characterisation of polymers used in coatings. Taguchi and co-workers [93] have developed a novel method for the direct analysis of small amounts of an oligomeric HALS occluded in PP material to study its photostabilising action on the basis of matrix-assisted laser desorption/ionisation mass spectrometry (MALDI-MS) using a solid sampling technique while avoiding troublesome solvent extraction. In this sampling protocol, the powdered mixture of PP composite sample containing trace amounts of an oligomeric HALS, Adekastab LA-68LD (MW = 1900), and the matrix reagent (dithranol) was spotted on the sample plate, then ion exchanged water was deposited onto the mixture to make a suspension, and finally, the dried mixture adhered on the plate was subjected to MALDIMS measurement. On the mass spectrum thus obtained by the solid sampling MALDI, the molecular ions of the HALS desorbed from the PP composite were clearly observed as three major series of the HALS components in the range up to about m/z 7000 with little interference by the PP substrate and the other additives. Moreover, in the MALDI-MS spectra for the UV-exposed sample, the satellite peaks around the major HALS components proved were enhanced significantly, reflecting the ability of the oxidised HALS species at the tetramethylpiperidine units to cause the photostabilising action. In addition, hydrolysed HALS species were also observed for the irradiated sample. These results suggest that not only the oxidation reaction but also the hydrolysis or decomposition of the oligomeric HALS components competitively proceeds in the PP composites during UV exposure. Figure 1.24 shows a typical MALDI mass spectrum of intact Adekastab LA-68LD measured by the conventional dried droplet method in linear mode along with the assigned structures of the major components. Table 1.6 summarises the calculated molar mass for the assigned structures and the observed peak top m/z values of the major components. Three series of the HALS components are mainly observed as the molecular ions (M+) on the mass spectrum in the range up to about m/z 8000. Here, the precise m/z values of b1 used as an internal standard for mass calibration were determined by an additional MALDI-MS measurement in reflector mode, confirming that the observed ions are mostly M+. Among these, the constituents designated with bn, which are the HALS molecules completely end capped with tetramethylpiperidyl groups, show the most intense peaks. In addition, the 45


21/3/07 11:21:39 am


Table 1.6 Calculated molar mass for the assigned structures and observed peak top m/z values of major components of HALS Calculated molar massa

Observed m/z value

a0

665.9

666.3

b0

791.1

791.5

c0

938.2

938.6

a1

1446.9

1447.1

b1

1572.1

1572.1b

c1

1719.2

1719.5

a2

2227.9

2228.2

b2

2353.1

2353.3

c2

2500.2

2500.6

a3

3008.9

3009.5

b3

3134.1

2124.5

c13

3281.2

3281.6

a4

3789.9

3789.9

b4

3915.1

3915.6

c4

4062.2

4063.0

a5

4570.9

4571.9

b5

4696.1

4696.7

c5

4843.2

4843.7

a6

5351.9

5352.9

b6

5477.1

5478.0

c6

5624.2

5623.8

a7

6132.9

6133.0

b7

3258.1

6258.3

c7

3405.2

6407.1

a8

6913.9

6915.1

b8

7039.1

7039.1b

c8

7186.2

7185.3

Component

a

: Isotopic abundance was taken into account : Used as internal standards for mass calibration: the molecular ions of b1, m/z 1572.1 and b8, m/z 7039.1 Source: Author’s own files b

46


21/3/07 11:21:39 am

Direct Determination of Additives in Polymers and Rubbers compounds containing a methoxy substituent (an) instead of a tetramethylpiperidyloxy unit of the corresponding main components (bn) and those with a spirochain-type terminals (cn) are also observed in fairly strong intensities. These two components are presumed to be the byproducts due to incomplete condensation or partial decomposition during synthesis of the HALS. The mass spectrum indicates that the oligomeric HALS consists

Figure 1.24 MALDI mass spectrum of Adekastab LA-68 LD obtained by conventional solution-based MALDI-MS. The weight ratio of sample/matrix was 1:10 Reproduced from Taguchi and co-workers, American Chemical Society [93]

47


21/3/07 11:21:39 am

Determination of Additives in Polymers and Rubbers of a number of components with at least three kinds of different chemical structures and wide variations of molecular weights. Figure 1.25 shows MALDI mass spectra of the HALS occluded in the PP composite (a) obtained by the dried droplet method after conventional solvent extraction from the PP

Figure 1.25 MALDI mass spectrum of HALS components in PP composite sample containing 1.0 wt% of HALS before UV irradiation: (a) solvent extracts from the sample obtained by solution-based preparation method; (b) HALS components directly desorbed from the PP composite obtained by the solid sampling technique Reproduced from Taguchi and co-workers, American Chemical Society [93]

48


21/3/07 11:21:40 am

Direct Determination of Additives in Polymers and Rubbers material and (b) directly desorbed from PP in the ionisation chamber by using the solid sampling technique. On the mass spectrum of the extracts (a), the peak of the main HALS oligomers markedly declined in comparison with those in Figure 1.24 so that the components in n = 6 and higher regions were scarcely observed. Moreover, the relative peak intensity of the byproduct a0 significantly increased and some satellite peaks around the major components such as b1 were fairly boosted or additionally observed after the solvent extraction probably due to the decomposition of the larger components. These results suggest that not only the insufficient extraction of the higher molecular weight HALS components but also their undesirable decomposition proceeded considerably during the extraction process. On the other hand, the mass spectrum obtained by direct MALDI-MS measurement of the PP sample (b) was almost identical to that of intact Adekastab LA-68LD shown in Figure 1.24. This fact suggests that the whole MW range of the HALS components was appropriately ionised during the solid sampling MALDI process through adequate contact between the matrix and the HALS molecules on the surface of the PP substrate. Here, the ions of the substrate polymer components and the antioxidants were scarcely observed on the mass spectrum under the given MALDI-MS conditions. These results demonstrate that MALDI-MS using the solid sampling method enables us to analyse the oligomeric HALS molecules occluded in the PP material directly without causing discriminative loss or decomposition of the HALS components during desorption. By using this technique, therefore, the subtle change in the molecular structure of the HALS components in PP during UV irradiation could be observed in the MALDI mass spectrum which it is possible to interpret in terms of the photostabilising action.

1.5 X-ray Photoelectron Spectroscopy (XPS) Using x-ray photoelectron spectroscopy, Pena and co-workers [94] examined the factors affecting the adsorption of organophosphorus polymer stabilisers on to carbon black.

1.6 Thermal Methods of Analysis

1.6.1 Differential Scanning Calorimetry Prasad and Shanker [95] used a differential scanning calorimeter (DSC) for the quantitative analysis of chemical blowing agents such as azodicarbonamide (azo) in commercial formulations. The DSC results were comparable to those obtained by the commonly used evolved gas analysis (EGA) technique. Advantages of DSC are: ease of operation,

49


21/3/07 11:21:40 am

Determination of Additives in Polymers and Rubbers shorter analysis time, environmental safety, and the quantitative analysis is independent of additives such as UV and antioxidant stabilisers which are normally present in carrier resins. The DSC technique is also effective in measuring azo concentrations up to 2% by weight, which is a limitation of the EGA technique. DSC can also be used to obtain the onset of the decomposition temperature and rate of decomposition of azo compounds containing Group II and Group IV metal salt activators, such as zinc oxide and zinc stearate. DSC also has the potential to detect the level of undecomposed blowing agent present in processed foam products as discussed next. DSC measures the temperature and heat flow associated with the transitions in materials as a function of time and temperature. Such measurements provide quantitative and qualitative information about the physical and chemical changes that involve exothermic and endothermic processes. Figure 1.26 is a general illustration of the type of information that DSC provides in the decomposition study of a foam concentrate. The melting of a carrier resin (in this case LDPE) is an endothermic process, whereas the decomposition of the azo is exothermic.

Figure 1.26 Typical DSC heating scans of azo dispersed in LDPE. The endothermic peak is due to the melting of LDPE. The exothermic peak is due to the decomposition of azo Reproduced with permission from A. Prasad and M. Shanker, Cellular Polymers, 1999, 18, 1, 35. ©Rapra Technology [95]

50


21/3/07 11:21:40 am

Direct Determination of Additives in Polymers and Rubbers The breadth of the exothermic curve (195 °C to 225 °C) indicates the temperature range over which the decomposition of azo occurs. The shape of the peak indicates the uniformity of the decomposition. In addition, the size of the peak, i.e., the area under the exothermic curve, is a quantitative measure of the amount of blowing agent that has decomposed in the sample. This integrated area under the exothermic peak is referred to as heat of decomposition, AHd. An exothermic peak at 230 °C followed by an endothermic peak at 250 °C was observed in the DSC heating scan of pure azo (scans are not shown here). The reason for the endothermic peak in pure azo is not known. This endothermic peak at 250 °C was not observed in any of the foam concentrates. Because of the overlap of the exothermic peak with the endothermic peak, it is difficult to obtain the heat of decomposition value of the pure azo compound. However, based on the measured heat of decomposition value of samples A-F, the heat of decomposition value of pure azo was estimated to be about 1200 J/g (see Table 1.7). Control samples (containing between 10% and 36% of blowing agent and between 90% and 64% of LDPE) were used to construct a calibration plot. These samples are free of any additives that would normally be present in the commercial samples. Some of the normalised DSC curves (total mass of 2.0 mg) as a function of % azo are shown in

Table 1.7 DSC results for the control foam concentrate samples Sample

Exothermic peak

Heat of decompositiona σ ΔHd

ΔHdb

Temperature (°C)

σ

A

222.5

0.6

120.5

5.0

1200

B

222.5

0.9

181.0

5.0

1206

C

221.8

0.9

241.0

6.0

1205

D

222.0

0.8

301.0

5.0

1204

E

220:0

0.9

361.0

3.0

1203

F

223.0

1.0

438.0

6.0

1216

100% Azo

230.5

1.5

-

-

-

(J/g of sample weight)

(J/g of azo)

σ: Refers to one standard deviation for n = 5 measurements a : ΔHd value obtained for the total sample weight of 2 mg, where % azo varies b : ΔHd value for 100% azo calculated from the known azo composition in samples A-F Reproduced with permission from A. Prasad and M. Shanker, Cellular Polymers, 1999, 18, 1, 35. ©Rapra Technology [95]

51


21/3/07 11:21:40 am

Determination of Additives in Polymers and Rubbers Figure 1.27. This figure clearly shows that the magnitude of the exothermic curve (area of the exothermic curve) increases with the azo concentration. This figure demonstrates that the relative heat of decomposition depends strongly on the amount of azo compound present in a given sample. Therefore, in principle, the DSC data can be used to construct a calibration plot of the azo concentration against ΔHd. Once a calibration plot has been established, routine samples run under the same conditions can simply be compared to the standard curves to measure the actual amount of azo compound present in unknown foam concentrate samples. Figure 1.28 shows a plot of ΔHd against percentage azo concentration for the control samples (A-F). The calibration plot of Figure 1.28 was constructed using a total sample weight of 2 mg. In Figure 1.28, data points represent the average value of five runs along with the 95% confidence limit error bars (2σn-1 for n = 5). The data points are represented by a straight line, which, as expected, passes through the origin. The following equation was generated using a linear least square fit to the data: ΔHd = 12.15* % azo Thus, % azo in an unknown sample can be determined by simply dividing the ΔHd, value by the slope of the calibration plot.

Figure 1.27 Normalised DSC scans showing the change in the exothermic peak area as a function of % azo. (a) sample B, (b) sample c, (c) sample E and (d) sample F Reproduced with permission from A. Prasad and M. Shanker, Cellular Polymers, 1999, 18, 1, 35. ©Rapra Technology [95]From Prasad and Shankar, Cellular Polymers [95]

52


21/3/07 11:21:41 am


Figure 1.28 Plot of % azo against heat of decomposition by DSC of samples (A-F) for a sample mass of 2 mg. Error bars represent the 2 x σn-1 limit for n = 5 measurements Reproduced with permission from A. Prasad and M. Shanker, Cellular Polymers, 1999, 18, 1, 35. ©Rapra Technology [95]

It has been shown that DSC is a useful, easy, fast and accurate method for the quantitative analysis of chemical blowing agents. The DSC results are comparable to the most commonly used EGA technique. It can quantitatively determine the amount of blowing agent present in both foam concentrates and finished foam sample. The most common types of additives, such as antioxidants and UV stabilisers, do not decompose in the temperature range of interest, thus making the quantitative determination of azo by DSC relatively easy. The system can be automated to reduce the operator’s time. DSC can also be used to study the effects of all formulation ingredients on the decomposition of blowing agent. Haldankar and Spencer [96] used DSC to investigate the thermal transitions occurring in polyacrylic acid and its sodium and potassium salts over a large range of water content at temperatures below the normal melting temperature of water. The bound water was identified as nonfreezing (type I), freezing with a constant melting temperature (type II), and freezing with a melting temperature dependent on the water content (type III). The

53


21/3/07 11:21:41 am

Determination of Additives in Polymers and Rubbers transition temperatures of the freezing states of the water were determined. Two constant melting temperatures were observed for the type II water in the sodium and potassium polyacrylates, while a single transition of this type was observed for polyacrylic acid. The sodium polyacrylate absorbed more water in the nonfreezing state than the potassium polyacrylate, and both polyelectrolytes absorbed about three times as much water in this state as the nonionic polyacrylic acid. The effects of water content on the occurrence of an exotherm at low temperature in the melting scans of the polyelectrolytes are described. A DuPont DSC/DTA 900 thermal analyser was used with a DSC cooling attachment. The DSC was purged with nitrogen and the subambient temperature was attained with liquid nitrogen. The cell constant was determined using a sapphire disc. The temperature scale of the DSC cell was calibrated using indium (mp = 156.6 °C), water (0 °C), cyclohexane (6.5 °C), and the crystallisation temperature of cyclohexane (−87.1 °C). With careful calibration and weighing, precisions of ±2% for the enthalpy of transition and ±1 °C for the transition temperature were obtained.

1.6.2 Differential Thermal Analysis Schwartz and co-workers [97] used isothermal differential thermal analysis to study the diffusion of Irganox 1330 (1,3,5 tris (3,5 di-tert-butyl-4-hydroxyl benzyl) mesitylene) in extruded sheets of isotactic polypropylene (iPP). Studies were conducted over the temperature range 80-120 °C. The measurements showed a clear relation between oxidation induction time and oxidation maximum time [both determined by isothermal dynamic thermal analysis (DTA)] and the concentration of stabiliser. It was possible to calculate the diffusion coefficients and the activation energy of diffusion of Irganox 1330 in iPP by measuring the oxidation maximum times across stacks of iPP sheets. For quantitative determination of the concentrations of antioxidants in PP that are required for the analysis of diffusion data, an isothermal DTA technique was developed that directly uses the effect of antioxidants on the thermooxidative stability of the polymers. Especially at elevated temperatures and in the presence of oxygen, polyolefines undergo thermooxidative degradation which follows a radical mechanism [98]. The time from the start of an isothermal DTA experiment to the beginning of exothermal decomposition is the so-called oxidation induction time (OIT). After this period, which depends on the antioxidant concentration, effectiveness, and temperature used, autocatalytic oxidation produces an exothermal peak [99-102]. The time from the start of the test to the maximum of this peak is the so-called oxidation maximum time (OMT) [103], which means the complete consumption of antioxidants and the loss of thermal stability of polymer. At elevated test temperatures, corresponding to short reaction times, it was difficult, or even impossible, to determine the OIT in the usual manner. For 54


21/3/07 11:21:41 am

Direct Determination of Additives in Polymers and Rubbers that reason OMT was chosen. The calibration curve, the OMT of iPP as a function of antioxidant concentration at 170 °C DTA temperature is shown in Figure 1.29. Each point in this figure is the average of 10 measurements carried out on iPP films with the specific antioxidant concentration. From the calibration curve it is obvious that at low levels of antioxidant concentration in the iPP film, the standard deviation is minimum. But at higher levels of antioxidant concentration, slight deviations are noted but they are within the acceptable level iPP film with 0.5% Irganox 1330, OMT = 2.50 ± 0.32 h; iPP film with 0.10% Irganox 1330, OMT = 5.77 ± 0.38 h). Sheets of iPP with 0.03 and 0.10% antioxidant levels were chosen to determine the influence of thermooxidative degradation during storage of the materials in the circulatingair oven. The plot of reciprocal temperature of the DTA oven versus the OMT for the unstabilised and stabilised iPP sheets is shown in Figure 1.30. For an iPP sheet with 0.03% antioxidant concentration, the OMT is about 2000 h at 120 °C. So the diffusion of the antioxidant in the iPP film can be measured at 120 °C for a period of 48 h without the influence of the thermooxidative degradation and the loss of added antioxidant. Films of iPP having the dimensions 15 mm x 15 mm x 100 µm with an antioxidant level of 0.03 or 0.10% were chosen for the diffusion measurements. Fifteen films having 0.03% antioxidant concentration were stacked and placed together and then placed over 15 films

Figure 1.29 oxidation maximum times for at 170 ºC of iPP sheets as a function of antioxidant concentration (isothermal DTA) Reproduced from Schwarz and co-workers, Journal of Applied Polymer Science [97]

55


21/3/07 11:21:41 am


Figure 1.30 Logarithms of oxidation maximum times of isocratic PP with different antioxidant concentrations as a function of reciprocal temperature (isothermal DTA) Reproduced from A. Prasad and M. Shanker, Cellular Polymers, 1999, 18, 1, 35. ©Rapra Technology [95]

with 0.10% antioxidant concentration, the whole stack of 30 sheets was kept tightly in the centre of diffusion device (two blocks of aluminium with steel bolts). This unit was placed in a circulating-air oven for several predeterminated time intervals at a constant temperature. Some experiments were performed at different isothermal conditions, namely, 80, 100, 110 and 120 °C. At the end of the run, the iPP sheets were separated and samples out of the centre of each sheet were analysed at 170 °C by isothermal DTA. This procedure is the measurement of residual stability time because the thermooxidative stability is determined after storage in the oven. After having stored a stack of iPP sheets for 48 h at 120 °C in a circulation-air oven, the residual stability time of each sheet of this stack was determined. Figure 1.31 shows the residual stability time at 170 °C (OIT and OMT) as a function of the thickness of the film stack. It is clear that both curves, OMT and OIT, are the same shape. Using the calibration curve shown in Figure 1.29 the concentration profile can be determined (Figure 1.32). Hatakeyama and co-workers [104] used differential thermal analysis and also DSC and TGA to determine the level of bound water in hydrophilic polymers. They were able to

56


21/3/07 11:21:42 am


Figure 1.31 Residual stability time (OMT and OIT) at 170 ºC of iPP sheets after storage of 48 h at 120 ºC as a function of the thickness of the film stack (isothermal DTA) Reproduced from A. Prasad and M. Shanker, Cellular Polymers, 1999, 18, 1, 35. ©Rapra Technology [95]

Figure 1.32 Antioxidant concentration of it iPP sheets after storage of 48 h at 120 ºC as a function of the thickness of the film stack (isothermal DTA) Reproduced from A. Prasad and M. Shanker, Cellular Polymers, 1999, 18, 1, 35. ©Rapra Technology [95]

57


21/3/07 11:21:42 am

Determination of Additives in Polymers and Rubbers distinguish between free water in the system and water bound to the polymer. Differential thermal analysis has been used to determine blowing agents in foams [105, 106].

1.6.3 Thermogravimetric Analyses This technique has been used to determine blowing agents in foamed plastics [105] to study liberation of stabilisers from PVC pipe at elevated temperature [107].

1.7 Vapour Phase Ultraviolet Spectroscopy Organic and inorganic pigments are used for coloration of polymers, polymer films and polymer coatings on metal containers. Vapour/phase UV absorption spectrometry at 200 nm has been used to identify such pigments [29]. In this method powdered samples are directly vaporised in the heated graphite atomiser. Thermal UV profiles of organic pigments show absorption bands between 300 and 900 °C, while profiles of inorganic pigments are characterised by absorption bands at temperatures above 900 °C. Temperature, relative intensity, and width of the bands allow the identification of the pigments. The technique shows fast acquisition of thermal UV profiles (2-3 minutes for each run), good repeatability and wide thermal range (from 150 to 2,300 °C). A 1:1 mixture of organic pigment yellow (2-nitro-p-toluidine coupled with acetoacetanilide) and inorganic PY 34 (lead chromate) was vaporised. The thermal UV profile clearly shows two absorption/bands at about 500 °C and 1,250 °C. The first band is attributed to the vapours which originate from the decomposition and pyrolysis of the organic pigment, the second band corresponds to the decomposition and vaporisation of lead chromate at high temperature (mp 844 °C). It is possible therefore to determine by a rapid run whether the pigment is a mixture or belongs to the organic or inorganic group.

1.8 X-Ray Fluorescence Analysis This technique has been applied to determining the identity of oxygen absorbers in polymers [108] also to determine traces of metals in polymers.

1.9 Nuclear Magnetic Resonance Spectroscopy The phase partitioning of additives in styrene-butadiene polymer blends has a large impact on the performance of the blend. Since solubility characteristics and processing of the blends influences partitioning, it is necessary to be able to quantify the level of Ionol

58


21/3/07 11:21:42 am

Direct Determination of Additives in Polymers and Rubbers (2,6 di-tert-butyl-4-methylphenol) in each phase. Smith and co-workers [109] have described an NMR method to quantify this partitioning based on the fact that the rubber phase and molecules dissolved therein, can be easily distinguished due to this phase’s enhanced motional characteristics. Table 1.8 and Figure 1.33 give the Ionol levels in the high-impact polystyrene (HIPS) as determined by liquid chromatography (LC) and the levels found in the rubber phase by NMR. The partition coefficient is defined as the ratio of the concentration of the Ionol found in the rubber phase to that found in the rigid PS phase. The level of Ionol in the rubber phase was determined by 1H NMR and the total amount in the HIPS was determined by LC. Since the level of rubber in the HIPS (9%) was also known, the concentration of Ionol in the PS phase could be calculated by difference. The ratio of these concentrations is the partition coefficient. Table 1.8 also lists the concentration of Ionol in the PS phase, calculated by subtracting the level in the polybutadiene from the total concentration in the HIPS, and the calculated partition coefficient for each sample. The average of these values is 2.0 and the estimated precision of these values is ±0.4. The value of 2.9 determined at low loadings of Ionol is probably due to the precision of determining Ionol by 1H NMR at that low loading. The partition coefficient is fairly constant at 1.6 to 1.9 for the samples containing from 3.5 to 7.3% Ionol, indicating the rubber phase is not being saturated.

Table 1.8 Weight% Ionol and partition coefficients in HIPS standards Nominol weight% Ionol in the HIPS

Weight% Ionol Weight% Ionol found in the found in the HIPS by LC rubber phase of the HIPS by NMR

Weight% Ionol in the PS phase by difference

Partition coefficient = Ionol concentration in PBD/Ionol concentration in PS

0.0

0.0

0.0

NA

NA

1.97

1.6

4.0

1.4

2.9

3.85

3.5

5.3

3.3

1.6

5.67

5.3

9.4

4.9

1.9

7.42

7.3

11.0

6.9

1.6

Reprinted with permission from P.B. Smith, W.C. Buzanowski, J.J. Gunderson, D.B. Priddy and L. Pfenninger in Proceedings of the 60th SPE Annual Technical Conference, ANTEC 2002, San Francisco, CA, USA, Paper No.421. ©2002, SPE [109]

59


21/3/07 11:21:42 am


Figure 1.33 The level of Ionel determined in the rubber phase by NMR and in the HIPS by LC, the slope giving the partition coefficient Reproduced from Smith and co-workers, SPE [109]

NMR spectroscopy has also been used in a limited number of other applications including the determination of stabilisers [110], water in polyols [111], starch in PE [77], degradation products of phosphorus containing additives [112], acrylic acid in oligomers [77], and plasticisers in PVC [112]. Wideline NMR spectroscopy has been used [113] for the determination of the plasticiser content, (e.g., di-iso-octyl phthalate) of PVC. The principle of the method is that the narrowline liquid-type NMR signal of the plasticiser is easily separated from the very broad signal that is due to the resin - integration of the narrow-line signal permits determination of the plasticiser. A Newport Quantity Analyser Mk I low-resolution instrument, equipped with a 40 ml sample assembly and digital readout, has been used to determine 20 to 50% of plasticiser in PVC. The sample may be in any physical state without significantly affecting the results, e.g., sheet samples are cut into strips 50 mm wide, which are rolled up and placed in the sample holder. A curvilinear relationship exists between the signal per g and the percentage by weight of the plasticiser. For highest precision, it is necessary to know the type of plasticiser present - use of the appropriate calibration graph gives a precision of ± 0.5%. However, one general calibration graph can be used. The precision is then approximately ±3%. As the NMR signal is temperature dependent, the temperature of calibration and of analysis should not differ by more than 4 °C. 60


21/3/07 11:21:43 am


References 1.

B. Vigerust, K. Kolset, S. Nordenson, A. Henricsen and K. Kleveland, Applied Spectroscopy, 1991, 45, 2, 173.

2.

Unscrambler: A Program for Multivariate Calibration and Prediction, Version 2.2. User’s Guide, CAMO, Trondheim, Norway, 1988.

3.

H. Martens and T. Naes, Multivariate Calibration, John Wiley and Sons, Chichester, UK, 1989.

4.

D.L. Massart, B.G.M. Vandeginste, S.N. Deming, Y. Michotte and L. Kaufman, Chemometrics: A Textbook, Elsevier, Amsterdam, The Netherlands, 1988.

5.

M.A. Sharaf, D.L. Illman and B.R. Kowalski, Chemometrics, John Wiley and Sons, New York, NY, USA, 1986.

6.

S. Wold and M. Sjostrom in Chemometrics: Theory and Application, Ed., B. Kowalski, ACS Symposium Series No.52, American Chemical Society, Washington, DC, USA, 1977, p.243-282.

7.

D.M. Haaland and E.V. Thomas, Analytical Chemistry, 1988, 60, 11, 1193.

8.

D.M. Haaland and E.V. Thomas, Analytical Chemistry, 1988, 60, 11, 1202.

9.

D.M. Haaland, Analytical Chemistry, 1988, 60, 11, 1208.

10. T. Kumar, Analyst, 1990, 115, 12, 1597. 11. R.W. Crecely and C.E. Day in Plastics Additives: An A-Z Reference, Ed., G. Pritchard, Chapman and Hall, London, UK, 1998, p.26-31. 12. S. Karlsson, Advances in Polymer Science, 2004, 169, 201. 13. S. Karlsson, F. Stangenberg and S. Agren, and Refined Routine Analysis for Quality Assessment of Recycled Plastics in Standardisation, KTH, Stockholm, Sweden, 2002. 14. W.L. Camacho, A. Vallés-Lluch and A. Ribes-Greus and S. Karlsson, Journal of Applied Polymer Science, 2003, 87, 13, 2165. 15. Y. Nishikawa, T. Nakano, H. Miyauchi, K. Nishikida and E.Y. Jiang, Applied Spectroscopy, 2004, 58, 958.

61


21/3/07 11:21:43 am

Determination of Additives in Polymers and Rubbers 16. F. Scholl, Atlas of Polymer and Plastics Analysis, Volume 3, Additives and Processing Aids, Spectra and Methods of Identification, 2nd Revised Edition, Carl Hanser Verlad, Munich, Germany, 1981, p.687. 17. F. De Thomas, J. Hall, D. Grzybowski in Leaping Ahead in Near Infrared Spectroscopy - Proceedings of the 6th International Conference on Near Infrared Spectroscopy, Eds., G.D. Batten, P.C. Flinn, L.A. Welsh and A.B. Blakeney, Australia, 1995, p.364-366. 18. J.W. Hall, D.E. Grzybowski and S.L. Monfre, Journal of Near Infrared Spectroscopy, 1993, 1, 1, 55. 19. S.P. Patel, Analytical Instrumentation, 1989, 44, 1331. 20. V. Di Noto, M. Saccon, B. Bresadola and R. Zannetti, Analyst, 1990, 115, 10, 41. 21. T. Kumar, Analyst, 1990, 115, 10, 1319. 22. M. Bowden, P. Donaldson, D.J. Gardiner, J. Birnie and D.L. Gerard, Analytical Chemistry, 1991, 63, 24, 2915. 23. T.L. Rapp, W.K. Kowalchyk, K.L. Davis, E.A. Todd, K.L. Liu and M.D. Morris, Analytical Chemistry, 1992, 64, 20, 2434. 24. D.W. Windig and J. Guilment, Analytical Chemistry, 1991, 63, 14, 1425. 25. K.C. Cole, C. Lehto and M. Propac in SPIE Optical Techniques for Industrial Inspection Conference, Quebec City, Canada, 1986, p.243-251. 26. R.J. Meilunas, J.G. Bentsen and A. Steinberg, Studies in Conservation, 1990, 35, 1, 33. 27. J. Soucek and E. Jelinkova, Analyst, 1982, 107, 1275, 623. 28. J. Lutzen, K. van Veen and S.T. Balke in Proceedings of the 60th SPE Annual Technical Conference, ANTEC 2000, San Francisco, CA, USA, Paper No.346. 29. P. Tittarelli, T. Zerlia, A. Colli and G. Ferrari, Analytical Chemistry, 1983, 55, 2, 220. 30. R.V. Albarino, Applied Spectroscopy, 1973, 27, 1, 47. 31. P. Hendra and H. Mould, International Laboratory, 1988, 18, 30.

62


21/3/07 11:21:43 am

Direct Determination of Additives in Polymers and Rubbers 32. F. Dywan, B. Hartmann, S. Klauer, M.D. Lechner, R.A. Rapp and M. Wöhlecke, Die Macromolekulare Chemie, 1993, 194, 6, 1527. 33. S. Ellahi, R.E. Hester and K.P.J. Williams, Spectrochimica Acta, 1995, 51A, 4, 549. 34. M.W. Hayes and A.G. Altenau, Rubber Age, 1970, 102, 5, 59. 35. R.P. Lattimer and R.E. Harris, Mass Spectrometry Reviews, 1985, 4, 3, 369. 36. R.P. Lattimer, Polymeric Materials, Science and Engineering, Spring Meeting, 2003, 88, 182. 37. R.P. Lattimer and K.R. Welch, Rubber Chemistry and Technology, 1980, 53, 1, 151. 38. R.P. Lattimer, R.E. Harris, D.B. Ross and H.E. Diem, Rubber Chemistry and Technology, 1984, 57, 5, 1013. 39. A.T. Jackson, A. Busy, J.H. Scrivens and K.R. Jennings in Proceedings of the 4th ASMS Conference on Mass Spectrometry and Allied Topics, Atlanta, GA, USA, 1981, p.750. 40. A.T. Jackson, A. Busy, K.R. Jennings and J.H. Scrivens, unpublished. 41. M. Barber, R.S. Bordoli, R.D. Sedgwick and A.N. Tyler, Journal of the Chemical Society, Chemical Communications, 1981, 7, 325. 42. N. Fürstenau, F. Hillenkamp and R. Nitsche, International Journal of Mass Spectrometry and Ion Processes, 1979, 31, 1-2, 85. 43. C.L. Johlman, C.L. Wilkins, J.D. Hogan, T.L. Donovan, D.A. Laude and M.J. Youssefi, Analytical Chemistry, 1990, 62, 11, 1167. 44. K. Peltonen, Analyst, 1986, 111, 7, 819. 45. P. Rudewicz and B. Munson, Analytical Chemistry, 1986, 58, 2, 358. 46. R.P. Lattimer, R.E. Harris, C.K. Rhee and H.R. Schulten, Analytical Chemistry, 1986, 58, 14, 3188. 47. I.V. Bletsos, D.M. Hercules, J.H. Magill, D. Van Leyen, E. Niehuis, A. Benninghoven, Analytical Chemistry, 1988, 60, 9, 938. 48. R.P.Lattimer, D.M. Harris, C.K. Rhee and H.R. Schulten, Rubber Chemistry and Technology, 1988, 61, 4, 639.

63


21/3/07 11:21:43 am

Determination of Additives in Polymers and Rubbers 49. L. O’Toole, R.D. Short, F.A. Bottino, G. Di Pasquale and A. Pollicino, Surface Interface Analysis, 1992, 18, 9, 667. 50. D.W. Carlson, M.W. Hayes, H.C. Ransaw, R.S. McFadden and A.G. Altenau, Analytical Chemistry, 1971, 43, 13, 1874. 51. A. Hilton and A. Altenau, Rubber Chemistry and Technology, 1973, 46, 4, 1035. 52. K.M. Baker, G. Chantrey and D. Mauer, Kautschuk und Gummi Kunststoffe, 1980, 33, 3, 175. 53. J. Pierre and J. Van Bree, Kunstoffe, 1983, 73, 6, 319. 54. R. Lattimer, Rubber Chemistry and Technology, 1988, 61, 4, 658. 55. R. Lattimer, H. Muenster and H. Budzikiewicz, Rubber Chemistry and Technology, 1990, 63, 2, 298. 56. H-R. Schulten, B. Plage and R. Lattimer, Rubber Chemistry and Technology, 1989, 62, 4, 698. 57. R. Lattimer and R. Harris, Rubber Chemistry and Technology, 1989, 62, 3, 548. 58. T. Yoshikawa, K. Ushimi, K. Kimura and M. Tamura, Journal of Applied Polymer Science, 1971, 15, 9, 2065. 59. A.W. Wims and S.J. Swarin, Journal of Applied Polymer Science, 1975, 19, 5, 1243. 60. R. Lattimer, Rubber Chemistry and Technology, 1995, 68, 5, 783. 61. S.J. Wright, M.J. Dale, P.R.R. Langridge-Smith, Q. Zhan and R. Zenobi, Analytical Chemistry, 1996, 68, 20, 3585. 62. A.P. Snyder, Journal of Analytical Applied Pyrolysis, 1990, 17, 127. 63. C.L. Johlman, C.L. Wilkins, J.D. Hogan, T.L. Donovan, D.A. Laude and M.J. Youssefi, Analytical Chemistry, 1990, 62, 11, 1167. 64. W.H. Waddell, K.A. Benzing, L.R. Evans, S.K. Mowdood, D.A. Weil, J.M. McMahon, R.H. Cody, Jr., and J.A. Kinsinger, Rubber Chemistry and Technology, 1991, 64, 4, 622. 65. Q. Zhan, R. Zanobi, S.J. Wright and P.R.R. Langridge-Smith, Macromolecules, 1996, 29, 7865.

64


21/3/07 11:21:44 am

Direct Determination of Additives in Polymers and Rubbers 66. K. Schwetlick, Pure and Applied Chemistry, 1983, 55, 10, 1629. 67. K. Schwetlick, T. König, C. Rüger, J. Pionteck and W.D. Habicher, Polymer Degradation and Stability, 1986, 15, 2, 97. 68. K. Schwetlick, J. Pionteck, T. König and W.D. Habicher, European Polymer Journal, 1987, 23, 5, 383. 69. D.G. Pobedimskii, V.I. Kurashov and P.A. Kirpichnikov, Journal of Polymer Science: Polymer Chemistry Edition, 1983, 21, 1, 55. 70. M.M. Tannahill, W.P. Enlow, Limitations of Liquid Chromatographic and FTIR Determinations for Ultranox 626 in Polyolefins, GE Specialty Chemicals Technical Bulletin, GE Plastics, 1988. 71. M. Reeder, W. Enlow and E. Borkowski in Proceedings of the SPE Conference Polyolefins VI, Houston, TX, USA, 1989, p.181. 72. X. Xiang, J. Dahlgren, W.P. Enlow and A.G. Marshall, Analytical Chemistry, 1992, 64, 22, 2862. 73. B. Asamoto, J.R. Young and R.J. Citerin, Analytical Chemistry, 1990, 62, 1, 61. 74. D.D. Bombick and J. Allison, Analytical Chemistry, 1987, 59, 3, 458. 75. W.J. Simonsick, Jr., private communication. 76. W.J. Simonsick, Jr., Progress in Organic Coatings, 1992, 20, 3-4, 411. 77. T. Tikuisis, D.E. Axelson and A. Shara, Polymer Engineering Science, 1993, 33, 1, 26. 78. R.J. Beuhler, E. Flanigan, L.F. Green and L. Friedman, Journal of the American Chemical Society, 1974, 96, 12, 3990. 79. F.W. Rollgen and H-R. Schulten, Organic Mass Specrometry, 1975, 10, 660. 80. D. Bombick, J.D. Pinkston and J. Allison, Analytical Chemistry, 1984, 56, 3, 396. 81. C.L. Johlman, C.L. Wilkins, J.D. Hogan, T.L. Donovan, D.A. Laude and M.J. Youssefi, Analytical Chemistry, 1990, 62, 11, 1167. 82. T. Chang, D. Waterman and M. Pignetti in Proceedings of the 40th ASMS Conference on Mass Spectrometry and Allied Topics, Washington, DC, USA, 1992, p.265.

65


21/3/07 11:21:44 am

Determination of Additives in Polymers and Rubbers 83. A.T. Jackson, K.R. Jennings and J.H. Scrivens, Rapid Communications in Mass Spectrometry, 1996, 10, 12, 1449. 84. A.T. Jackson, A. Buzy, K.R. Jennings and J.H. Scrivens, European Journal of Mass Specrometry, 1996, 2, 2, 115. 85. C.R. Shick, Jr., P.A. DePalma, Jr., and R.K. Marcus, Analytical Chemistry, 1996, 68, 13, 2113. 86. L. O’Toole, R.D. Short, F.A. Bottino, A. Pollucino and A. Recca, Polymer Degradation and Stability, 1992, 38, 2, 147. 87. J.J. Kirkland, C.H. Dilks and S.W. Rementer, Analytical Chemistry, 1992, 64, 11, 1296. 88. I.V. Bletsos, D.M. Hercules, D. Van Leyen, B. Hagenhoft, and E. Niehuis and A. Benninghoven, Analytical Chemistry, 1991, 63, 18, 1953. 89. S.G. McKay and P.J. Pachuta, Polymer Preprints, 1996, 37, 299. 90. A.S. Helton and A.G. Altenau, Rubber Chemistry and Technology, 1973, 46, 1035. 91. V. Svob and F. Flajsmon, Croatian Chimica Acta, 1970, 42, 417. 92. S.D. Hanton, JCT Coatings Technology, 2004, 1, 1, 62. 93. Y. Taguchi, Y. Ishida, H. Ohtani and H. Matsubara, Analytical Chemistry, 2004, 76, 3, 697. 94. J.M. Pena, N.S. Allen, M. Edge, C.M. Liauw and B. Valange, Polymer Degradation and Stability, 2001, 72, 1, 31. 95. A. Prasad and M. Shanker, Cellular Polymers, 1999, 18, 1, 35. 96. G.S. Haldankar and H.G. Spencer, Journal of Applied Polymer Science, 1989, 37, 11, 3137. 97. T. Schwarz, G. Steiner and J. Koppelmann, Journal of Applied Polymer Science, 1989, 38, 1, 1. 98. R. Gächter and H. Müller, Kunststoff-Additive, 2nd Edition, Carl Hanser Verlag Munich, Germany, 1993.

66


21/3/07 11:21:44 am

Direct Determination of Additives in Polymers and Rubbers 99. E. Kramer and J. Koppelmann, Kunstoffe, 1983, 73, 11, 714. 98. E. Kramer, Osterreichische Kunstoffe Zeitschrift, 1984, 15, 3-4, 32. 99. E. Kramer and J. Koppelmann, Polymer Degradation and Stability, 1986, 14, 4, 333. 100. E. Kramer and J. Koppelmann, Polymer Degradation and Stability, 1986, 16, 3, 261. 101. E. Kramer and J. Koppelmann, Polymer Engineering and Science, 1987, 27, 13, 945. 102. G. Steiner and J. Koppelmann, Kunstoffe, 1987, 77, 3, 286. 103. G. Steiner and J. Koppelmann, Polymer Degradation and Stability, 1987, 19, 4, 307. 104. T. Hatakeyama, K. Nakamura and H. Hatakyama, Thermochimica Acta, 1988, 123, 1, 153. 105. R.H. Hensen, SPE Journal, 1962, 18, 77. 106. R.H. Hensen, SPE Journal, 1962, 18, 79. 107. W. Stachn and H. Sendel, Labor Praxis, 1988, 12, 308. 108. J. Lopez-Cervantes, D.I. Sanchez-Machado, S. Pastorelli, R. Rijk and P. PaseiroLosada, Food Additives and Contaminants, 2003, 20, 3, 291. 109. P.B. Smith, W.C. Buzanowski, J.J. Gunderson, D.B. Priddy and L. Pfenninger in Proceedings of the 60th SPE Annual Technical Conference, ANTEC 2000, San Francisco, CA, USA, Paper No.421. 110. W. Barendswaard, J. Moone and M. Neilen, Analytica Chimica Acta, 1993, 283, 3, 1007. 111. R.B. Ray, C. Kradgel and, L. McDermott, Polymer Materials Science and Engineering, 1988, 58, 542. 112. D.F. Slonaker and D.C. Sievers, Analytical Chemistry, 1964, 36, 6, 1130. 113. P.B. Mansfield, Chemische Industrie, 1971, 28, 792.

67


21/3/07 11:21:44 am


68


21/3/07 11:21:45 am

Extraction Techniques for Additives in Polymers

2

Extraction Techniques for Additives in Polymers Author

2.1 Introduction Most polymers contain quite complex additive systems which are incorporated during manufacture to impart beneficial properties during manufacturing operations, e.g., protective antioxidants, slip agents, etc., and during end-use, e.g., antioxidants, ultraviolet (UV) stabilisers, plasticisers and antistatic agents, flame retardants, antioxidants and thermal stabilisers. The first stage in the examination of a polymer, either from the point of view of identifying the polymer or identifying and determining additives present must be to separate gross polymer from gross additives. It may then be necessary to separate the gross additive fraction into individual additives by a chromatographic procedure in order to facilitate the identification of individual additives. Separation of gross polymer from gross additives is necessary so that examination of the polymer is not affected by interference from the additives present and also vice versa. Many different procedures have been studied for the removal of gross additives from the polymer and some of these are discussed next. The main extraction procedures used are summarised in Table 2.1 and although most of them can be carried out on material cut to a particle size of less than 2 mm diameter, it is often advantageous to produce material of a smaller size and with a larger surface area to mass ratio. This is conveniently done by grinding at the temperature of liquid nitrogen using an efficient and easily cleaned cutter mill. The extraction of additives which are strongly adsorbed or chemisorbed onto the polymer filler matrix must be carefully watched by the analyst, as a change of the method of manufacture of, for example, the filler or in the method of compounding the plastic formulation, can also markedly alter the degree of adsorption being produced and hence invalidate an established quantitative extraction procedure. The use of ‘stronger’ extraction reagents can cause complications at the measurement stage and hence each system must be carefully screened and frequently checked. Polymer extraction procedures using organic solvents do not extract all types of organic additives from the polymers, also many inorganic compounds and metal inorganic compounds, e.g., calcium stearate, are 69


21/3/07 11:21:45 am


Table 2.1 Examples of extraction methods Extraction method

Polymer or compound

Solvent

Additives or contamination extacted

Single solvent (Soxhlet)

Polyvinyl chloride (PVC)

Diethyl ether

Plasticisers

Single solvent (reflux)

Polyethylene

Chloroform - 1,1,1trichloroethane

Antioxidants

Polypropylene

Ethylene dichloride - trichloroacetic acid

Chemisorbed amides/ amines

Polyester

Methanol - Karl Fischer reagent

Water

Polyolefines

Toluene - methanol, xylene - methanol

UV absorbers, antioxidants, slip agents

Acrylics

Acetone - light petroleum

Plasticisers, lubricants

Packaging films

Water - diethyl ether

Odour and taintforming additives

Solvent + reagent (reflux)

Solution/ precipitation

Steam/solvent distillation Vacuum/thermal extraction

Nylon fluorocarbon polymers

Water, process fumes

Source: Author’s own files

insoluble. The presence of metals will have been indicated in the preliminary examination of the polymer. However, most types of organic polymer additives can be readily extracted from polymers with organic solvents of various types. The first stage is to solvent-extract the total additives from the polymer in high yield and with minimum contamination by low molecular polymer. Extracts should be used for analysis without delay as they may contain light or oxygen sensitive compounds. When delay is unavoidable, storage in actinic glassware under nitrogen in a refrigerator minimises the risk of decomposition. In practice, the most commonly used procedures for the separation of gross polymer and gross additives are based on fractional precipitation and fractional extraction. In the fractional precipitation procedure a good solvent for the polymer is required, also a nonsolvent for the polymer in which the additives remain soluble. Information on solvents and non-solvents for various polymers is given in Table 2.1. It must be realised that the solubility of a polymer depends not only on the type of polymer but also the degree of polymerisation, i.e., molecular weight and degree of branching and crosslinking and on its steric configuration. Thus a relatively low molecular weight unbranched polyethylene (PE) may have a high solubility in toluene whilst a high molecular weight highly branched

70


21/3/07 11:21:45 am

Extraction Techniques for Additives in Polymers PE would have a low solubility even in hot toluene. In the case of styrene-butadiene copolymer, the uncrosslinked polymer is soluble in aromatic solvents, whilst the highly crosslinked (gel) fraction is completely insoluble and, indeed, this can be used as the basis of a method for separating gel from uncrosslinked polymer. Copolymers usually dissolve in a greater number of solvents than homopolymers. Thus, whilst PVC is only slightly soluble in acetone or methylene chloride, its copolymers with vinyl acetate or acrylates dissolve easily.

2.2 Solvent Extraction

2.2.1 Polyolefins The scheme outlined in Figure 2.1 has been proposed for the solvent extraction and examination of PE [1]. Some useful solvents are listed in Table 2.2.

Figure 2.1 Analysis of polyolefins [1]

71


21/3/07 11:21:45 am


Table 2.2 Solvent extraction methods of additive extraction from polymers Polymer type

Substance extracted

Extracting solvent

Comments

Ref.

Type of extraction

PE

Cresolic and phenolic antioxidants

Chloroform

Heat at 50 °C for 3 hours in a closed container

[1]

Fractional extraction

PE

Cresolic antioxidants

Hexane

Heat at 50 °C for 24 hours

[2]


PE

Cresolic antioxidants

Ether

In the dark at 20 °C

[3]


PE

Phenolic antioxidants

Chloroform

[4]


PE

Antioxidants

Toluene

Reflux to dissolve polymer in precipitate with methanol

[5]

Fractional precipitation

PE

Antioxidants

Water

At 70 °C under nitrogen

[6]



Details of particular solvent extraction schemes for polyethylene are described in more detail next.

• Toluene-Methanol Extraction System Total internal plus external additives can be extracted from low and high-density polyethylene (HDPE) and polystyrene (PS) by procedures involving solution or dispersion of the polymer powder or granules (3 g) in cold redistilled sulfur-free toluene (50-100 ml) followed in the case of PE, by refluxing for several hours. Rubber-modified PS does not completely dissolve in toluene if it contains gel.

• Methylene Dichloride Extraction Methylene dichloride is a particularly good solvent for PP extraction because of its high volatility. In addition to additives, most solvents also extract some low molecular weight polymer with subsequent contamination of the extract. To overcome this, Slonaker and Sievers [2] have described a procedure for obtaining polymer-free additive extracts from PE based on low temperature extraction with n-hexane at 0 °C. This procedure is also applicable to PP and PS. 72


21/3/07 11:21:45 am


• Chloroform or 1,1,1-trichloromethane Extraction A portion of the PP sample (2-20 g) is weighed accurately into a 250 ml round-bottomed flask. To this is added 50 ml of chloroform (or 1,1,1-trichloromethane), which is then placed in the flask under a water-filled reflux condenser. The mixture is gently brought to the boil on a water bath and heated for 30 minutes. The solution is cooled and then decanted into a 100 ml volumetric flask after filtering through a No. 42 Whatman filter paper. Any polymer collected in the filter paper is transferred to the round-bottomed flask and 40 ml of chloroform is added. The extraction procedure is then repeated. The cooled solution is filtered into the volumetric flask containing the first extract. The residue is washed with a little more chloroform, filtered into the volumetric flask and diluted to the 100 ml mark with chloroform. The solution is shaken well. Alternatively, the sample can be ground in a Wiley Mill to pass through a 60 mesh screen. A 10-20 g portion of sample is extracted in a Soxhlet apparatus with heptane for a minimum of 6 hours. The extract is transferred to a 300 ml tall-form beaker, and evaporated to approximately 20 ml, and then diluted to 40 ml with chloroform.

• Ethanol-Methanol Extraction System The sample to be analysed must be very thinly sheeted or powdered. A sample (1.000 ± 0.0005 g) is wrapped with an extraction cloth, which has been previously treated to remove sizing, and so on. The sample is placed in an Underwriter’s extraction cup and extracted for 16 hours with 95% ethanol or methanol. The alcoholic extract is transferred to a 100 ml volumetric flask, cooled to room temperature and diluted to the mark with the extraction solvent. Alternatively, the polymer compound containing the unknown stabiliser is cut or ground into small pieces, and about 5 grams is extracted by heating under reflux for 24 hours with 50 ml of ethanol. The extract is cooled to room temperature and the polymer is filtered off.

• Cyclohexane Extraction A representative sample of PE is ground in the Apex Mill. The sample (1 g) is placed into a 100 ml round-bottomed flask and cyclohexane (20 ml) is added. A condenser is fitted to the flask and the cyclohexane is allowed to reflux on a water bath for 1 hour. The condenser is washed with 20 ml cyclohexane, the flask is removed from the bath, and then cooled to room temperature and shaken well. After filtering through a No. 802 filter paper into a 100 ml volumetric flask, the filtrate is diluted to the mark. 73


21/3/07 11:21:46 am


• Diethyl Ether Extraction Polypropylene (2-5 g) is extracted with diethyl ether for 8 hours in a Soxhlet continuous extraction apparatus. The weight to be used will vary with the type and amount of additive expected. If these are not known take the maximum amount, if possible. The diethyl ether is removed by distillation on a water bath and ethanol (5 ml) is added to each flask. The flasks are left to stand for 10 minutes. If necessary, at the end of this time, the resulting solutions are filtered quantitatively to remove any precipitated polymer. If required, the contents of the flask are evaporated to dryness and the contents are then dissolved in a suitable solvent for further analysis of additives. The extracted polymer is left to air dry. Alternatively, polymer (5 g) is weighed into an extraction thimble, the diethylether (130 ml) is added, followed by extraction for 3 hours. The system is cooled to room temperature and the solvent is drained into a 250 ml flat bottom flask. The extraction thimble and extraction apparatus are rinsed with 25-30 ml of extraction solvent and drained into a 250 ml flask (a turbid solution due to the precipitation of extracted polyolefin may be seen upon cooling, however, this is not uncommon nor should any special precautions be taken). The extraction solvent is removed by evaporation to leave a dry residue.

• Heptane Extraction Samples of PP sheets (50 x 50 x 0.5 mm thick ) are weighed and then refluxed in 50 ml of boiling heptane. The efficiency of extraction is better than 98% after 3 hours of extraction.

• Acetonitrile Extraction The sample is cut into small shavings with a drill bit prior to extraction. The plastic shavings (1 g) are extracted overnight with 5 ml of acetonitrile. The extraction is performed at ambient temperature in a sealed amber vial with constant stirring. The additives are sufficiently soluble in the acetonitrile extracting solvent due to the small amounts of analyte present. The extract solutions are filtered prior to analysis.

• Decalin Extraction Extraction with decalin for 30 ml at 110 °C has been used to extract antioxidants from PE and PP [7, 8]. 74


21/3/07 11:21:46 am


2.2.2 Polystyrene

• Toluene Extraction Sample (5.0 + 0.01 g) is weighed and transferred to a 250 ml Pyrex glass centrifuge bottle. Toluene (50 ml) is added into the bottle. A polythene coated magnetic stirrer rotor is dropped into the bottle which is stoppered with a cork (avoid rubber bungs). The bottle is placed on a magnetic stirrer and left for several hours until the sample has completely dissolved or dispersed in the solvent. Absolute alcohol (50 ml) is accurately pipetted into the gently stirred contents of the bottle to precipitate the polymer. The resulting solution is centrifuged to isolate the solvent phase from the polymer phase.

• Methyl Ethyl Ketone Methyl ethyl ketone or propylene oxide, are alternative suitable solvents for PS. Dissolved polymer is reprecipitated by the addition of methyl alcohol or absolute ethanol (up to 300 ml) and the polymer is removed by filtration or centrifugation. The additive-containing extract is then be gently concentrated to dryness as described previously.

• Diethyl Ether Extraction Sample (5 ± 0.1 g) is weighed and then transferred to a 50 ml beaker containing 25 ml diethyl ether. The beaker is covered with a watch glass and then left overnight. The ether is decanted carefully into a large beaker and the beads are washed with two further portions (10 ml) of ether, decanting each washing into the main extract. Using an air line and a warm water bath the ether is carefully evaporated off until about 1 ml of solution remains. The solution is transferred by a 1 ml pipette to a 2 ml volumetric flask. The walls of the beaker are rinsed with 5-10 ml of ether and concentrated to about 0.5 ml. This solution is transferred to the 2 ml volumetric flask and the flask contents are diluted to 2 ml with diethylether.

• Ethanol or Methanol Extraction The sample to be analysed must be very thinly sheeted (or powdered by passing through a Wiley Mill). A sample (2.00 + 0.02 g) is accurately weighed and wrapped with an extraction cloth which has been previously extracted to remove sizing, and so on. The sample is placed in an Underwriter’s extraction cup and extracted for 16 hours with 95% ethanol 75


21/3/07 11:21:46 am

Determination of Additives in Polymers and Rubbers or methanol. The alcohol extract is transferred to a 100 ml volumetric flask, cooled to room temperature and diluted to the mark with the extraction solvent.

2.2.3 Acrylic Polymers A fractional reprecipitation procedure involving a solution of the polymer in acetone, followed by reprecipitation with light petroleum can be used to separate additives from acrylics. The acetone - light petroleum extract can be used for the determination of plasticisers and lubricants (Figure 2.2).

2.2.4 PVC Diethyl ether is a favoured solvent for removing additives from this polymer. It has been used to extract stabilisers, lubricants and plasticisers from PVC. Figure 2.3 shows a scheme involving ether extraction for the separation of additives from PVC [4-6, 9-10]. Table 2.3 gives the results of following four hours of extraction with ether by extraction with other solvents. It can be seen that extraction for four hours with carbon tetrachloride/methanol azeotrope gives better results than extraction for 15 hours with methanol.

Figure 2.2 Analysis of acrylic samples. Source: Author’s own files

76


21/3/07 11:21:46 am


Figure 2.3 Analysis of PVC compositions. Source: Author’s own files

77


21/3/07 11:21:46 am


Table 2.3 Multiple extraction Plasticiser

Concentration (%)

Extracts (%) 4 hour ether, 5 hour MeOH

4 hour ether, 4 hour CCl4, MeOH

Bisoflex 791

28.5

27.7

28.7

Tritolyl phosphate

28.5

25.8

28.9

Mesamoll

28.5

25.5

28.1

Bisoflex 795

28.5

28.1

29.0

Reoplex 220

28.5

20.4

29.3

Hexaplas PPA

23.5

11.0

17.4

Reproduced with permission from J. Haslam and D.C.M. Squirrell, Analyst, 1955, 80, 957, 871 [14]

Haslam and Soppet [11] found that extraction of PVC with acetone, followed by precipitation of dissolved polymer with light petroleum, gave poor results - only 28.5% was recovered from a composition containing 31.8% tritolyl phosphate. Substitution of 1, 2-dichloroethane for acetone gave no improvement, but diethyl ether extracted 31.8% of the sample, and the extract contained a negligible amount of PVC. For routine extraction it was recommended by Haslam and Soppet that the sample be stood overnight in cold ether, then extracted for 6-7 hours in the apparatus. This procedure has also been described by Döhring [12], who specified that anhydrous ether should be used. Thinius [13] compared the rates of extraction of dioctyl phthalate by ether, carbon tetrachloride and petroleum, at room temperature. In 70 minutes the ether extract from a composition containing 40% plasticiser amounted to 39.7%. In the same time the carbon tetrachloride rate of extraction was 33.1% and the petroleum extract was 29.9%. Extraction with carbon tetrachloride for 64 hours gave only 35.6% extract. Thinius also found that mixtures of phthalate and phosphate esters could be completely removed by ether, but light petroleum gave only 65% of the expected yield. Toluene dissolved some PVC at room temperature, and very much more in a Soxhlet extraction. For compositions containing polypropylene adipate, which is only partly extracted by ether, Haslam and Squirrell [14] used a 6 hour ether extraction, followed by an 18 hour methanol extraction. The ether extract was 34.5% from a composition containing 36.1% of a mixture of equal parts phthalate and tritolyl phosphate, but only 33.1% from a composition containing 45.7% of a mixture of equal parts of dioctyl phthalate, tritolyl phosphate and polypropylene adipate.

78


21/3/07 11:21:47 am

Extraction Techniques for Additives in Polymers The combined ether and methanol extracts amounted to 34.7% and 44.3%, respectively. Wake [15] quotes the results of some unpublished work carried out at the laboratories of RAPRA - methanol extracted 40.5% and 36.8% from compositions containing 42.3% polypropylene sebacate and 36.4% polypropylene adipate, respectively. The extracts contained PVC equivalent to 0.6% and 0.9%, respectively. Clarke and Bazill [16] extracted with ether for 15 hours then with methanol for 8 hours. They state that ether removes plasticisers whose molecular weight is less than 1000. Korn and Woggon [17] used methanol or diethyl ether to extract diphenylthiourea and 2-phenylindole dicyanamide from PVC. The extraction of plasticisers from PVC compositions must involve diffusion processes, and the rate of extraction should depend, to some extent, on the initial plasticiser concentration. This was found to be true for the extraction of Mesamoll by ether [18]. Compositions made by ICI, Corvic D65/1, lead stearate (0.1%), and Mesamoll (10-50%) were extracted with ether for 1-8 hours, and the plasticiser contents m (% w/w), at times t (hours), were calculated from the weights of the extracts. Figure 2.4 shows the curves obtained by plotting log (m/mo) against t, for different values of mo, the initial plasticiser content. In every case, except when mo = 10%, there is initially a very rapid loss of plasticiser, followed by a period when m/mo decreases exponentially. When m/mo = 10%, the rate of decrease of m/mo is very small at all stages of the extraction. An approximate solution of the diffusion equation is: dc/dt = div (D grad c) This equation is appropriate to diffusion out of isotropic solids of arbitrary shape when D is constant. In this case the diffusion equation reduces to: dc/dt = DV2c and the proportion of the diffusing substance remaining in the solid after time t is given by: c - cs = α exp (-β2 (S/V)2Dt) co - cs β

(2.1)

where co is the concentration in the solid at t = 0, cs is the surface concentration at t > 0, c is the average concentration in the solid at t > 0, α and β are shape constants, and S/V is the surface-to-volume ratio. It has been shown that it is applicable in all cases if (S/V)2Dt is greater than 0.6. The samples used in the extraction experiments, performed by Robertson and Rowley [18], had a surface-to-volume ratio of about 50 cm-1, so that for t > 1 hour and D > 10-7 cm2/s, equation (2.1) should be applicable to their results, if D is constant.

79


21/3/07 11:21:47 am


Figure 2.4 Effect of extraction time (t) and plasticiser concentration (mo) on ether extraction of Mesamoll from PVC Reproduced from Robertson and Rowley, British Plastics [18]

Assuming that os = 0, S/V is constant, and m/mo = c/c0 (none of these assumptions can be exactly true at all stages of the extraction, but should be nearly correct during the later stages), Equation (2.1) can be written as: log (m/mo) = A – βt

(2.2)

Where β is proportional to (S/V)Dt. It can be seen that after the first few hours the curves are of the form given in Equation (2.2), which can therefore, be assumed to described adequately the important later stages of the extraction process. In the earlier stages the departures from linearity are probably due to the rapidity with which D decreases as m decreases. Other solvent extraction procedures for the isolation of plasticisers in PVC are reviewed in Table 2.4. 80


21/3/07 11:21:47 am


Table 2.4 Plasticisers in PVC Solvent Extraction Method

Ref.

Diocytylphthalate tritolylphthalate

Diethyl ether extraction, weighing

[11, 12]

Plasticisers

Solvent extraction, weighing

[18]

Polypropylene adipate Dioctylphthalate Tritolylphthalate

Diethyl ether then methanol extraction, weighing

[14]

Polypropylene sebacate Polypropylene adipate

Methanol extraction

[15]

Gas Chromatography Plasticisers

Solvent extraction, gas-liquid chromatography (GLC)

[19, 20]

Alkyl adipate Alkyl phthalate

Solvent extraction, GLC

[21, 22]

Benzylbutyl phthalates


[23]

Alkyl phthalate Fatty acid esters


[24, 25]

Dimethyl phthalate Dimethyl sebacate Triacetin Diacetin Diethyl phthalate

Dichloromethane extraction, GLC

[26]

Adipate and phthalate type

Mild pyrolysis - gas chromatography (Py-GC)

[27]

Plasticisers

Solvent extraction, gas chromatography (GC)

[28]

Plasticisers

Hydrolysis to alcohols, GC

[29]

Dibutyl phthalate Di-isobutyl phthalate Bis (2-ethylhexyl) phthalate Di-n-octyl phthalate

Py-GC

Alkyl phthalate Alkyl adipates Alkyl azealates Alkyl citrates Alkyl sebacates Alkyl glycollates Alkyl phosphates

Tetrahydrofuran extraction, GLC (includes determination of alcohol degradation products)

[32]

Plasticisers

Py-GC of polymer

[33]

[30, 31]

81


21/3/07 11:21:47 am


2.2.5 Rubbers Various extraction solvents for rubbers are listed in Table 2.5.

Table 2.5 Extraction solvents from rubber Extraction solvent

Extraction procedure

Rubbers Amine and phenolic antioxidants

Type of additive

Ethanol/HCl

Reflux, then steam distill [34] amines from extract

Ref.

Rubbers Phenyl salicylate, resorcinol benzoate

Ether

[35]

Rubbers Antioxidants

Acetone

[36]

Acetone Rubbers Ketone-amine condensates, phenols, 2-mercaptobenzimidazole

[37]


2.2.6 Polyacrylamide A weighed quantity of polymer (1-10 g) is added to 50 ml of a methanol/water solution (80:20 v/v) in a glass bottle. A magnetic stirring bar coated with Teflon is added and the solution is stirred to extract the acrylamide monomer for at least 3 hours, preferably overnight.

2.2.7 Polyurethane A weighed sample of polyurethane foam is covered with 75 ml of methanol in a 250 ml beaker and soaked for 5 minutes with occasional compression. The methanol is decanted, squeezing the foam as completely as possible to express the methanol. This is repeated twice with fresh methanol and the combined extracts are concentrated to 25 ml.

2.2.8 Vinyl Chloride, Butadiene, Acrylonitrile, Styrene, 2 Ethylhexyl Acrylate Copolymers A weighed portion of the copolymer is placed in a septum vial containing measured amounts of N, N-dimethylacetamide. The vials are heated to 90 °C to aid dissolution of the polymer. When the dissolution is complete, the vials are cooled to room temperature.

82


21/3/07 11:21:48 am

Extraction Techniques for Additives in Polymers A portion of distilled water is forcibly injected into each polymer solution. The vials are shaken briefly to assure complete mixing of the water with the organic phase and to prevent the precipitated polymer from forming a film on top of the solution. Special care and specialist techniques are required when dealing with laminates and surfacecoated films. For major components the separation is made quantitatively and the analysis is completed by various techniques. For volatile components, separation, identification and quantification can often be carried out in one analytical process.

2.2.9 Other Polymers Steam-solvent distillation using diethyl ether has been used to remove and analyse for odour and taint from additives in food packaging films. Another technique that has been used is vacuum/thermal extraction. This procedure has been applied to polyamides and fluorocarbon polymers. The procedure is used for the direct isolation or release of volatile components from a polymeric matrix and may involve the combined use of vacuum and heat, as for example in the mass spectrometer direct insertion probe or during dry vacuum distillation. Alternatively, the volatiles may be swept from the heated sample by a flow of inert gas for concentration by freeze trapping and/or collection on to a solid adsorbent prior to thermal or solvent desorption for GC or mass spectrometric (MS) examination.

2.3 Fractional Precipitation In this method a solution or dispersion of a plastic is stirred into at least 10 times its volume of a solvent, which acts as a precipitant for the dissolved plastic but which dissolves completely in the first solvent. The precipitated plastic is pulverised and repeatedly extracted with the precipitant/solvent or re-dissolved and precipitated. Or a non-solvent is added to a solution of the plastic until the first faint turbidities appear due to the high polymer plastic components. The solvent is then reduced on a water bath or by vacuum distillation. The solvent and the non-solvent must be completely miscible and the solvent must have a lower boiling point than the non-solvent. Or the plastic is dissolved in a solvent, which dissolves the plastic at higher temperatures - the plastic separates out on cooling while the additives remain in solution. Or plastic dispersions can often be separated by stopping the action of the emulsifier and/or dispersing agent, i.e., breaking the emulsion. The following methods can be used according to the type of emulsifier:

83


21/3/07 11:21:48 am

Determination of Additives in Polymers and Rubbers 1. Alteration of the solvent phase. 2. Precipitation of the emulsifier by the addition of acids or a interfacially active counterion. 3. Freezing the dispersion out with solid carbon dioxide. 4. Subjection of the dispersion to dialysis, whereby the water-soluble inorganic salts and relatively low molecular emulsifiers are diffused through cellophane membranes so that the protective colloid and the polymer can now be easily separated quantitatively by centrifugal action. This method is labour intensive but very effective and if it does not completely release any chemisorbed constituents from the polymer - filler matrix it will often leave them in a form very vulnerable to attack by the analytical reagent(s) finally used in the determination.

2.4 Fractional Extraction Fractional extraction can be used in three ways, depending on the substance under examination: (a) Finely shredded solid plastics are extracted one-by-one in a Soxhlet apparatus with solvents, of increasing solvent power. (b) To ensure that the plastic particles are fine enough to allow the components to be extracted to diffuse sufficiently quickly, a solution or dispersion of the plastic is poured over an inert carrier, such as silica gel or Sterchamol, the solvent is allowed to evaporate and the carrier is then extracted in a heatable column with solvents of increasing solvent power. (c) The components of a plastic are distributed between two non-miscible liquids in separating funnels or a distribution apparatus. Most of these procedures can be carried out on material cut to a particle size of less than 2 mm diameter - it is often advantageous to produce material of a smaller size and with a large surface area to mass ratio. This is conveniently done by grinding at the temperature of liquid nitrogen using an efficient and easily cleaned cutter mill. The extraction of additives strongly adsorbed or chemisorbed on the polymer - filler matrix must be carefully observed by the analyst, as a change of the method of manufacture of, for example, the filler or in the method of compounding the plastic formulation, can

84


21/3/07 11:21:48 am

Extraction Techniques for Additives in Polymers also markedly alter the degree of adsorption being produced and hence invalidate an established quantitative extraction procedure. The use of ‘stronger’ extraction reagents can cause complications at the measurement stage and hence each system must be carefully screened and frequently checked. In a modification of this procedure the polymer is refluxed with a reagent, which decomposes additives present to put them in a form, which is soluble in the solvent. Thus when PP on which is chemisorbed fatty acid amides or amines, is refluxed with a solution of ethylene dichloride and trichloroacetic acid then the amides are decomposed as follows: RCONH2R´ + H2O = R´NH2

2.5 Separation by Diffusion Methods The plastic is precipitated from a solution or dispersion in a glass vessel as a thin film on the inner wall; meanwhile a solvent, which is a non-solvent towards the plastic, is added in several stages. The solvent must be able to dissolve the additives. Additives or a second plastic can be isolated in a short time by diffusion alone if the mixture is left to stand in the solvent, occasionally shaken and the solvent is renewed four to six times.

2.6 Dialysis or Electrodialysis This method has only been used for separating low-molecular weight salts and interfacially active substances from protective colloids and high polymer plastics and for fractionating pure plastics according to their molecular weight. Some examples of these four separation procedures are shown in Table 2.6.

2.7 Vacuum Thermal Displacement Extraction Method These procedures are used extensively for the direct isolation or release of volatile components from a polymeric matrix and may involve the combined use of vacuum and heat, as for example in the MS direct insertion probe or during dry vacuum distillation. Alternatively, the volatiles may be swept from the heated sample by a flow of inert gas for concentration by freeze trapping and/or collection on to a solid adsorbent prior to thermal or solvent desorption for GC or MS examination.

85


21/3/07 11:21:48 am


Table 2.6 Examples of separating procedures Plastic

Operation

Aim or separating procedure

(a) Fractional precipitation procedures (precipitation by addition of non-solvent for polymer to solvent solution of polymer). Polyvinyl chloride

Stirring of a concentrated solution Plasticisers, emulsifiers remain of tetrahydrofuran into methanol in solution, PVC and its polymers are precipitated

Polyacrylate

Stirring the solution in acetone into 20 times the amount of water

Emulsifiers, water soluble resins and salts remain in solution, polymerisate is precipitated

(b) Fractional precipitation (titration of solvent solution of polymer with non-solvent for polymer). Polyacrylate

Water is added to the solution in Emulsifiers, water soluble resins acetone until faint turbidity occurs, and salts remain in solution, acetone is then distilled off under polymerisate is precipitated vacuum

(c) Fractional precipitation (thermal precipitation of polymer from solution at low temperature). Polyethylene

Dissolve in enough benzene to make a clear solution appear when warmed then cooled

Polymerisate is precipitated, additives soluble in cold benzene, paraffin, waxes, resins remain in solution

(d) Emulsion breaking methods (1) Addition of solvent Polymethacrylatedispersions

Dispersion is stirred into 20 times its amount of methanol or isopropanol

Polymerisate is precipitated, emulsifier and any protective colloid remain in solution

(2) Emulsifier precipitation by acid addition Rubber or butadiene copolymers

Dispersion is broken by acidifying After acidifying, fatty or resin as long as soap is used as the acid is extractable with ether; emulsifier latex is precipitated

(3) Freezing out dispersion by Cardice addition Polyvinyl acetate or silicone oil dispersions

Action of emulsifier destroyed by cold

Polymerisate is precipitated, can now be filtered off or centrifuged off

(4) Application of dialysis Polyvinyl acetate ester or polyacrylate dispersions

Salts and low molecular emulsifiers are dialysed

In the dialysate, salts and emulsifiers, in the dialysate residue, protective colloid polymerisate

86


21/3/07 11:21:48 am


Table 2.6 Continued Plastic

Operation

Aim or separating procedure

Fractional extraction procedures Solvent extraction of shredded plastic Polyvinyl chloride

Consecutive extractions with CCl4, ether, benzene, CH2CCl2, tetrahydrofuran

Isolation of stabilisers and plasticisers, recognition of copolymers or poly blends

Solution or dispersion of plastic passed over inert carrier on which plastic adsorbs, solvent evaporation from carrier, then desorbtion from carrier with strong solvents Polyethylene

Consecutive extraction of silica with gasoline, methanol, CHC13, water, methanol or acetone, then benzene or toluene (perhaps warm)

In the gasoline: waxes; in CH3OH: emulsifiers; in CHC13: montan waxes in water: cellulose derivatives and polyvinyl alcohol; in benzene/ toluene: polyethylene

Plastics components distributed between two non-miscible liquids Polyvinyl acetate

Extracted from aqueous methanol Isolation of plasticisers solutions or dispersions with pentane/ether mixtures

Separation by diffusion Polyvinyl acetate

Polymer precipitated with Ligroin or ligroin/ether mixtures

Isolation of the plasticisers, can also be used

Separation by dialysis or electrodialysis Polyacrylate

See (4)


2.7.1 Effects of Polymer Milling on Extraction Schröder [38] warns that during solvent extraction techniques there may be complications, which can result in faulty interpretation of results due to stabiliser rearrangements or decompositions. However, although oxidation of the polymer additives during extraction may be avoidable, there exists the danger that crushing the polymer prior to extraction may lead to a sequence of reactions, which affect the chemical structure of the inhibitor. All polymers of longer chain length suffer from degradation when under mechanical stress which usually starts by a chain rupture in the centre of the macromolecule and results in the formation of macro radicals. This degradation is accelerated at low temperatures and even occurs when cutting with a knife. Pazonyi and co-workers [39] found a linear relationship between the radical concentration (determined by the consumption of diphenyl-picryl87


21/3/07 11:21:49 am

Determination of Additives in Polymers and Rubbers hydrazyl) and the surface (Figure 2.5) when cutting PE and plasticised PVC. They could prove that this could occur when the chemical bonds with polymers, which to some extent are present in the elastic state are cut by mechanical forces with 100% efficiency. In the presence of inhibitors reactions between macro radical and inhibitor may occur even in the absence of oxygen. As a consequence reaction products with the polymers themselves, but also deactivation products of the intermediately formed inhibitor radical are to be expected. Workers at the National Bureau of Standards accept that an attachment of butyl residues of dibutyl-tin-diacetate to PVC radicals occurs in the presence of UV radiation: R + (C4H9)2Sn (CH3COO)2 → R – C4H9 + C4H9Sn (CH3COO)2 which has been confirmed by examination of results obtained by Frye and co-workers [40] with 14C butyl-labelled organotin compounds. In the presence of oxygen the probability of reaction of the macro radicals with the inhibitor system is increased considerably if phenolic or amine antioxidants are present. Further losses on reaction induced decomposition products mainly occur when the extracts obtained are re-concentrated. Even the inherent volatility of some antioxidants - above all that of the phenols and aromatic amines (Table 2.7) is so high that it provides the basis of a direct determination in the polymer by vacuum sublimation as suggested by Yushkevichyute

Figure 2.5 Consumption of diphenyl picrylhydrazyl radicals in relation to surface area of polyethylene and PVC Reproduced from Schröder, Pure and Applied Chemistry [38]

88


21/3/07 11:21:49 am


Table 2.7 Volatility of antioxidants [40] Antioxidant

Vapour pressure (mm Hg)

Loss of weight (% at 150 °C)

2,6-Di-t-butyl-p-cresol

22.15

100

2-Benzyl-6-t-butyl-p-cresol

1.83

100

2,2′-Methylene-bis-6-t-butyl-p-cresol

0.169

19 ... 28

Diphenylamine

7.52

100

N-Iso-propyl-N′-phenyl-p-phenylene-diamine

0.59

40 ... 53

N-N′-Diphenyl-p-phenylenediamine

0.032

2 ... 3


[6] for the determination of antioxidants in PE and can also be applied to the explanation of the mechanism of action of phenolic and amine antioxidants in PE and PP. Thus a separation by distillation of the 2,6-di-t-butyl-4-methylphenol from its dimer deactivation product at 100 °C was successful and provided evidence for the isomerisation of the phenoxy radicals formed primarily to oxybenzyl radicals and their recombination to dioxydiphenylethane:

On the other hand the high volatility of additives occurring in the classic method of reconcentrating polymer extracts by distillation - even by evaporation of the solutions will result in considerable loss of substance as was found by Schröder [38] in the case of 2,6di-t-butyl-4-methylphenol. He found an evaporation loss of 0.75% after 24 hours when storing Ionol with a large surface in stagnant air. This loss increased to 63% when the chloroform solution was evaporated in a fume cupboard. Schröder [38] also points out that the distillation processes - even careful freeze-drying should be avoided for quantitative work in systems of such high volatility as the stabilisers,

89


21/3/07 11:21:49 am

Determination of Additives in Polymers and Rubbers especially if the quantity and type of the decomposition products are also of interest. In such cases only an enrichment by chromatographic processes should be considered. From the point of view of separation from the polymer, gel-permeation chromatography is a possible method since the separation of molecules in the pores of the gels is mainly achieved according to particle size, and the large polymer molecules can be excluded from the separation process by suitable selection of the pore size distribution of the gels so that they leave the separation column together with the solvent. Such a separation of low molecular substances from polymer mixtures have been described for plasticisers [41], oils [42] and methylsilanols [43]. It is to be seen from the work of Coupek and co-workers [44] and others [45, 46] who deal systematically with this problem, as to how far the stabiliser mixture is separated subsequently. The effects of mechanical degradation by polymer crushing on stabiliser structure, such as those discussed previously, are, of course, avoided in separation methods based on dissolving the polymer in a solvent, then precipitating the polymer, but not the stabiliser, with a nonsolvent, providing a solvent extract which contains only the stabiliser. Again, however, this process needs a consideration of the solution-precipitant effects on the stability, especially of the reaction products of stabilisers or their fragments with the polymer. Such reaction products have been both determined and isolated with PVC, PE, PP and natural rubber. Thus, Frye and co-workers [47] were able to provide evidence by infrared (IR) spectroscopy [46] and radiochemistry [48] of the insertion of ester groupings from barium, cadmium or zinc-carboxylates into PVC after heat treatment. Tin contents and organic residues of organotin stabilisers in heat-treated PVC are indicated by the same authors [47, 49]. Schlimper [50] and Schröder and co-workers [51] proved the direct reaction between PVC and nitrogen-containing organic stabilisers by elemental analytical examinations and infrared and UV spectrometry. Phenolic antioxidants or their decomposition products in part were re-found in PP after oxidative degradation. With rubber vulcanisates hydrochloric acid-resistant amine-rubber compounds have been reported after thermal oxidation in the presence of aromatic amines [52].

2.8 Solvent Extraction – Infrared Spectrometry Udris [53] has described two schemes, based on chromatography and infrared spectroscopy, for the identification, in PVC extracts of: i) dialkyltin dialkylthioglycollates: R SCH2 COOR´ Sn R SCH2 COOR´´

90


21/3/07 11:21:49 am

Extraction Techniques for Additives in Polymers dialkyltin dilauryl mercaptides: R SCH2 (CH2)10 CH3 Sn R SCH2 (CH2)10 CH3 R, R´, R´´ = alkyl stabilisers in which the dialkyltin group is combined with both thiol and carboxyl groups. ii) dialkyltin maleates R OO CCH Sn R OO CCH dialkyltin dialkyl maleates R OO CCH = CH COOR´ Sn R OO CCH = CH COOR´´ dialkyltin dilaurates R OOC (CH2)10 CH3 Sn R OOC (CH2)10 CH3 It is first of all necessary to prepare an extract of the PVC in which the organotin stabiliser is quantitatively recovered. For the extraction of the first set of compounds, i) the polymer is refluxed with acetone and the silver salts precipitated by the addition of aqueous silver nitrate. After drying the residue is examined by infrared spectroscopy to identify alcohols, thioacids, thiols, and alkyl groups attached to tin. For the extraction of the second set of compounds ii) the polymer is refluxed with 10% aqueous sodium hydroxide prior to the identification of alcohols by gas chromatography then the alkyl groups attached to tin and carboxylic acids are identified. Udris [53] examined methods for the recovery and identification of tin stabilisers in excess plasticiser (90 to 95%). The plasticisers included diisooctyl phthalate and mixtures of diisooctyl phthalate and tritolyl phosphate in various proportions. Good samples of dialkyltin dichlorides were isolated by this method from mixtures of tin stabilisers with diisooctyl phthalate and tritolyl phosphate; the dialkyltin groups have been identified in dibutyl tin dinonylmaleate, dioctyltin dilaurate, dibutyltin dinonylthioglycollate and dioctyltin thioglycollate.

91


21/3/07 11:21:50 am

Determination of Additives in Polymers and Rubbers Renier and co-workers [54] give details of a method for the infrared spectral analysis of Methacrol 2138 (polydiisopropylaminoethylmethacrylate-co-decylmethacrylate) and Santowhite (4,4´ butylidene bis (6-tert-butyl-m-cresol) in extracts of polyethyl urethane urea (PEUU) biomedical implantable material. The extraction was carried out in vitro using methanol. Figure 2.6 a-c provides a quantitive comparison of the infrared spectra of polyether urethane urea and Santowhite. The extract spectrum differed from that of the bulk of the polymer by the presence of bands from Santowhite in the 1630-1150 cm-1 region and by the increased intensity and breadth of the 3324 cm-1 (ν (N-H)) and 1711 cm-1 ν (C=O) hydrogen bonded bands of the urethane group. Table 2.8 gives the IR bands in the extract of polyether urethane urea that are attributed to the extracted Methacrol 2138 and Santowhite. Quantitative determination of Santowhite in the extracts was made with the band at 1386 cm-1, a δ(C-H) methylene band of Santowhite and a normalising band νs(CH2 – O) at 2797 cm-1 which was assigned to the ether functionality of the PEUU. Patticini [55] has described an IR method for the determination of 1 to 8% of mineral oil in PS. In this method the PS sample is dissolved in carbon tetrachloride, together with known mineral oil standards. The solutions are evaluated by measurements made between 3,100 and 3,000 cm-1 using a spectral subtraction technique.

Table 2.8 New major FT-IR bands in polyether urethane urea extracts Bands (cm-1) Assignmenta

Extract B

C

D

2998

νA(CH3)

SW

-

SW

1730

ν (C=O)

-

MT

MT

1598

ν (C=C) aromatic

SW

-

SW

1490-1358

δ + ω(C-H) aliphatic + δ(CCH3) and δ(O-H)

SW

-

SW

1413

δ(C=C) aromatic

SW

-

SW

1272

ν(C-O) + δ(O-H)

SW

-

SW

1035

ν(C-O) or δ(C-H) aromatic

SW

-

SW

a: ν stretch; δ: bend; ω: wag Reproduced with permission from M. Renier, J.M. Anderson, A. Hiltner, G.A. Lodoen and C.R. Payert, Journal of Biomaterials Science – Polymer Edition, 1993, 5, 3, 231 [54]

92


21/3/07 11:21:50 am


Figure 2.6 Qualitative comparisons of the infrared spectra. (a) PEUU B, (b) B extract, and (c) Santowhite powder showing the influence of Santowhite powder on the B extract spectrum in the 1630-1150 cm-1 region and the increased intensities of the υ(N-H) band at 3324 cm-1 and the υ(C=O) hydrogen bonded carbonyl band at 711 cm-1 from the PEUU urethane group Reproduced with permission from M. Renier, J.M. Anderson, A. Hiltner, G.A. Lodoen and C.R. Payet, Journal of Biomaterials Science – Polymer Edition, 1993, 5, 3, 231 [54]

Supercritical fluid chromatography (SFC) coupled with Fourier transform infrared spectroscopy has been used to determine polymeric surfactants in various polymers [56, 57]. Spell and Eddy [58] have described IR spectroscopic procedures for the determination of up to 500 ppm of various additives in PE pellets following solvent extraction of additives at room temperature. They showed that Ionol (2,6-di-t-butyl-p-cresol) and Santonox R

93


21/3/07 11:21:50 am

Determination of Additives in Polymers and Rubbers (4,4´-thio-bis-(6-t-butyl-m-cresol) are extracted quantitatively from PE pellets by carbon disulfide in 2-3 hours and by iso-octane in 50-75 hours. The carbon disulfide extract is suitable for scanning in the infrared region between 7.8 and 9.3 nm, whilst the iso-octane extract is suitable for scanning between 250 and 350 nm. Robertson and Rowley [18] studied the extraction of plasticisers from PVC using different solvents prior to analysis by IR spectroscopy. Table 2.9 gives the results of following 4 hour extractions with ether, by extractions with other solvents. It can be seen that extraction for 4 hours with carbon tetrachloride/methanol azeotrope gives better results than extraction for 15 hours with methanol. If a composition containing a polyester plasticiser and a monomeric plasticiser is extracted first with ether, then with methanol or carbon tetrachloride/methanol, the ether extract

Table 2.9 Multiple extractions Plasticiser

Concentration, %

Extracts, % 4 h ether, 5 h MeOH 4 h ether, 4 h CCl4MeOH Bisoflex 791 28.5 27.7 28.7 Tritolyl phosphate 28.5 25.8 28.9 Mesamoll 28.5 25.5 28.1 Bisoflex 79S 28.5 28.1 29.0 Reoplex 220 28.5 20.4 29.3 Hexaplas PPA 23.5 11.0 17.4 Reproduced with permission from M.W. Robertson and R.M. Rowley, British Plastics, 1960, 33, 1, ?? [18]

Table 2.10 Extraction of mixed plasticisers Solvent Time (h) Extract (%) Diethyl ether 8 28.6 Carbon tetrachloride 4 33.4 Diethyl ether 4 27.3 27.8 Methanol 15 0.5 Diethyl ether 4 27.2 33.2 Carbon tetrachloride/methanol 4 6 Reproduced with permission from M.W. Robertson and R.M. Rowley, British Plastics, 1960, 33, 1, ?? [18]

94


21/3/07 11:21:50 am

Extraction Techniques for Additives in Polymers will contain nearly all the monomeric plasticiser, and some of the polyester, and the second extract will contain almost pure polyester. Table 2.10 gives the results of extracting a composition containing Reoplex 220 (19.0%), Bisoflex 791 (7.5%), and Bisoflex 79S (7.0%), using various procedures, and Figure 2.7 shows the IR spectra of the extracts and the original plasticisers.

Figure 2.7 IR spectra of PVC plasticisers and extracts (a) mixture of Reoplex 220 (5607%), Bisoflex 791 (22.4%), (b) Reoplex 220, (c) carbon tetrachloride extract, (d) ether extract, (e) carbon tetrachloride-methanol extract Reproduced from Robertson and Rowley, British Plastics [18]

95


21/3/07 11:21:51 am


2.9 Solvent Extraction – Ultraviolet Spectroscopy Straightforward UV spectroscopy is liable to be in error because of interference by other highly absorbing impurities that may be present in the sample [59-62]. Interference by such impurities in direct UV spectroscopy has been overcome or minimised by selective solvent extraction or by chromatography [60]. However, within prescribed limits UV spectroscopy is of use and, as an example, procedures are described next for the determination of Ionol (2,6-di-tert-butyl-p-cresol) and of Santonox R (4,4´-thio-bis-6-tert-butyl-m-cresol) in polyolefins [63-66]. Certain additives, e.g., calcium stearate and thiodipropionate, do not interfere in the determination. Other phenolic antioxidants, e.g., Ionox 330, Topanol CA and Santonox R, do interfere.

2.9.1 Ionol in Polyolefins A polymer sample (20 g) was weighed into a 250 ml round-bottomed flask and 50 ml of chloroform was added. A water-cooled condenser was connected and the flask was placed on the electric heating mantle and brought gently to boiling point. The solution was refluxed for 30 minutes at a moderate rate. After cooling, the chloroform solution was carefully decanted into a 100 ml volumetric flask (using a filter if necessary) and stoppered immediately. A further 40 ml of chloroform was added to the contents of the round-bottomed flask and a second extraction of the polymer was carried out. When cool, the contents were filtered into the volumetric flask containing the first extract. The residue was washed with sufficient chloroform to dilute it to the mark. The flask was shaken to ensure homogeneity. The UV spectrum of the chloroform extract against a chloroform blank was recorded from 250 to 310 nm using 1 cm cells. The absorbance of the Ionol absorption peak was measured at 278 nm (see Figure 2.8). The concentration of Ionol CP present in the extract was determined by reference to the prepared calibration curve.

2.9.2 Santonox R In Polyolefins A representative sample was ground in the Apex Mill. About 1 gram of the sample was weighed into a 100 ml round-bottomed flask and cyclohexane (20 ml) was added. A condenser was fitted to the flask and the cyclohexane was allowed to reflux on a water bath for 1 hour. The condenser was washed with 20 ml cyclohexane, the flask was removed from the bath, cooled to room temperature and then shaken well. The solution was filtered through a No. 802 filter paper into a 100 ml separating funnel. The filter

96


21/3/07 11:21:51 am

Extraction Techniques for Additives in Polymers paper was washed with a further 10 ml cyclohexane. Freshly prepared sodium hydroxide solution (25 ml) was added and then the solution was shaken for 3 minutes then allowed to settle. The caustic layer was run into a 100 ml standard flask. The extraction was repeated with another 2 x 25 ml portions of alkali. The extract was made using 1 cm silica cells and sodium hydroxide solution as a blank. The base line absorbance stated in the calibration procedure was calculated as follows and the concentrations of Santonox R read from the graph: ABL = A266 – A335 – 0.684 (A236 – A335) where ABL = base line absorbance, A335, A266, A236, = absorbance at 335, 266, 236 nm, respectively. The base line absorbance was plotted against the concentration of Santonox R in mg/100 ml. Ruddle and Wilson [66] have described a UV spectroscopic method for the characterisation of phenolic stabilisers in solvent extracts of polymer compositions.

Figure 2.8 Determination of Ionel CP in polyolefins by UV spectroscopy. Source: Author’s own files

97


21/3/07 11:21:51 am

No change Reproduced with permission from L.H. Ruddle and J.R. Wilson, Analyst, 1969, 94, 1115, 105 [66]

x8 336, 304 277 1,3,5-Trimethyl-2,4,6 tris-(3,5-di-t-butyl-4hydroxy-benzyl) benzene Ionox 330

303, 274

578 x30 428 277 Bis-(3,5-di-t-butyl-4-hydroxyphenyl)methyl Binox M

303, 225

Absorption supressed x8 340, 286 303, 274, 257 277 2,6-Di-t-butyl-4-methylphenol Topanol OC

Alkaline reaction product λmax (nm) Absorption change after nickel peroxide reaction Nickel peroxide reaction product λmax (nm) Alkaline ethanolic solution λmax (nm) Ethanolic solution λmax (nm) Trade name Chemical constitution

Table 2.11 Wavelengths of maximum absorption in ethanol, ethanol-potassium hydroxide, ethanolnickel peroxide and ethanol-nickel peroxide – potassium hydroxide

Determination of Additives in Polymers and Rubbers These workers point out that usually the additive must be separated in a pure state from co-extracted additives usually by thin-layer chromatography (TLC) and then identified by measurement of the UV, IR, nuclear magnetic resonance (NMR) and mass spectra of the compound. This full treatment is required only for new stabilisers - for a characterisation of well known compounds the simplest method is by direct comparison of the UV absorption spectra with those of a series of known stabilisers. For some compounds this will probably be sufficient, but many substituted phenols have similar spectra, and for three of the most frequently used antioxidants the UV spectra are identical. Topanol OC, Ionox 330 and Binox M (see Table 2.11 for their chemical constitution) in ethanolic solution all have λmax = 277 nm, with a shoulder at 282 nm. To extend this procedure Ruddle andWilson [66] prepared the spectra of alkaline solutions of the phenols, which were then measured either directly against a solvent blank or as ‘difference spectra’ measured against the neutral solution. This still gives almost identical spectra for the three compounds mentioned previously. Ruddle and Wilson [66] developed two further stages in this procedure for extending the use of UV spectrophotometry in the characterisation of these compounds. They consist of (a) measuring the UV absorption spectrum of the stabiliser solution after reaction with solid nickel peroxide and (b) remeasuring it after making the reaction products alkaline. Cook [67] obtained the substituted stilbene quinone after

98


21/3/07 11:21:51 am

Extraction Techniques for Additives in Polymers reaction of 2,6-di-t-butyl-4-methylphenol with lead dioxide, and Braithwaite and Penketh [68] have used lead dioxide for the determination of Topanol OC in liquid paraffin. They obtained an absorption peak with λmax = 420 nm. Stafford [69] used air for the oxidation and obtained a product with λmax = 365 nm. Kharasch and Joshi [70] have reported on the oxidation of bis-(3,5-di-t-butyl-4-hydroxyphenol) methane, carried out by pumping oxygen into an alkaline ethanolic solution of the phenol. They obtained a dark purple solution produced by the anion of

The reaction at room temperature of dilute solutions of antioxidant (about 1 mg/100 ml), with nickel peroxide is complete within 2 minutes and, if continued for longer periods, the absorbance of the strongest band at 286 nm begins to diminish. The sets of spectra for Topanol OC, Binox M and Ionox 330, are shown in Figure 2.9 (a, b and c). Table 2.11 gives the wavelengths of maximum absorption for each procedure, and also some indication of the absorbance change, i.e., the dilution required, after nickel peroxide reaction. The initial concentration of the stabilisers was 10 mg/100 ml of ethanol. In an attempt to overcome the difficulty of interference effects by other polymer additives in the UV spectroscopic determination of phenolic antioxidants, Wexler [71] makes use of the bathochromic shift exhibited by phenols on changing from a neutral or acidic medium to an alkaline one. This shift is due to the change of absorbing species because of solutesolvent interaction. Using a double-beam recording spectrophotometer, he measured a difference spectrum by placing an alkaline solution of the polymer extract in the sample beam, and an identical concentration of sample in acid solution in the reference beam. The resulting difference spectrum is a characteristic and useful indication of the concentration and chemical identity of the phenolic substance. Possible interferences due to non-ionising, non-phenolic species are usually cancelled out in the difference spectrum which should make the technique of interest to the polymer analyst. Typical spectra obtained for an antioxidant are shown in Figure 2.10. These spectra exhibit two maxima and two minima. Close adherence to Beer’s law is usually obeyed by the difference peak spectra. Soucek and Jelinkova [72] have used differential UV spectra run in acid and alkaline solutions to determine 2,6-di-tert-butyl-4-methylphenol and 4 substituted 2,6-xylenol in PP. These two substances have virtually identical spectra in the absence of alkali. In alkaline medium 4 substituted 2,6-xylenol forms a phenolate readily, thus allowing use of the bathochromic shift for its determination. 99


21/3/07 11:21:52 am


Figure 2.9 UV spectra for (a) Topanol CA, (b) Binox M, and (c) Ionox 330. Curve A: ethanolic solution, curve B: alkaline ethanolic solution (10 ml of ethanolic solution plus 1 ml of water, plus 2 ml of ethanolic potassium hydroxide solution), curve C: ethanolic solution after reaction with nickel peroxide, and curve D: nickel peroxide reaction product made alkaline with two drops of ethanolic potassium hydroxide solution Reproduced from Kharash and Joshi, Journal of Organic Chemistry [70]

100


21/3/07 11:21:52 am


Figure 2.10 UV spectra of 4,4´-thiobis-(6-tert-butyl-m-cresol) exhibiting bathochromic shift in alkaline medium Reproduced from Wexler, ACS [71]

Scheele and co-workers [73, 74] have found extensive agreement between conductiometric and UV spectroscopic methods of quantitative antioxidant analysis. Crompton [75] in a method for the determination of Nonox CI (N,N-di-β-napththyl-pphenylenediamine) in HDPE describes a 1½ hour toluene extraction of the polymer under reflux, performed under a nitrogen blanket and mentions that, under these conditions no oxidation of the antioxidant occurs. In this way he was able to distinguish between true Nonox CI oxidation occurring during polymer manufacture and ‘accidental’ oxidation occurring during the preliminary solvent extraction stage of analysis. He also mentions that, under these conditions, some 5% of the additive remains unextracted in the polymer, but that this is allowed for in the method of calibration which involves refluxing known quantities of Nonox CI with virgin PE, as in the sample extraction procedure.

2.9.3 Styrene Monomer Crompton and co-authors [76] have described a UV spectroscopy method for the determination of styrene monomer in chloroform.

101


21/3/07 11:21:52 am

Determination of Additives in Polymers and Rubbers In this method the PS sample (0.5 g) is dissolved in 50 ml of chloroform or another suitable spectroscopic solvent and the UV spectrum recorded in the region 280-310 nm against polymer-free solvent in the reference cell. If the polymer is incompletely soluble in the solvent (e.g., due to the presence of gel or pigments), it is dissolved in 30 ml of solvent and the suspension filtered rapidly under light vacuum through an asbestos pad to remove insolubles. Conventional and high-impact polystyrene (HIPS) contains various non-polymer additives (e.g., lubricants) which result in widely different and unknown background absorptions at the wavelength maximum at which styrene monomer is evaluated (292 nm). The influence of the background absorptions on the evaluation of the optical density due to styrene monomer is overcome by the use of an appropriate baseline technique, claimed to make the method virtually independent of absorptions due to polymer additives. In this technique a straight line is drawn on the recorded spectrum across the absorption peak at 292 nm in such a way that the baseline is tangential to the absorption curve at a point close to the absorption minima occurring at 288 nm and 295-300 nm (Figure 2.11). A vertical line is drawn from the tip of the styrene absorption peak at 292 nm to intersect the baseline and the height of this line is then a measure of the optical density due to the true styrene monomer content of the test solution. This baseline correction technique can obviously be applied to the determination of styrene monomer in PS only if any other UV absorbing constituents in the polymer extract (e.g., lubricant, antioxidants) absorb linearly in the wavelength range 288-300 nm. If the polymer extract contains polymer constituents other than styrene with non-linear absorptions in this region, then incorrect styrene monomer contents will be obtained. An obvious technique for removing such non-volatile UV absorbing compounds is by distillation of the extract followed by UV spectroscopic analysis of the distillate for styrene monomer as described next. The PS sample (0.5 g) is dissolved in chloroform or ethylene dichloride (20 ml) in a stoppered flask and the solution is poured into an excess of methyl alcohol (110 ml) to reprecipitate dissolved polymer. The polymer is filtered off and washed with methanol (120 ml) and the combined filtrate and washings gently distilled to provide 200 ml of distillate containing only styrene monomer and any other distillable component of the original polystyrene sample. Non-volatile polymer components (namely stabilisers, lubricants and low molecular weight polymer) remain in the distillation residue. The optical density of the distillate is measured either at 292 nm or by the baseline method against the distillate obtained in a polymer-free blank distillation. Calibration is performed by applying the distillation procedure to solutions of known weights of pure styrene monomer in the appropriate quantities of methyl alcohol and the chlorinated solvent. Table 2.12 shows results obtained for styrene monomer determinations carried out on two samples of pigmented polystyrene by the direct ultraviolet method and by the distillation

102


21/3/07 11:21:52 am


Figure 2.11 Typical UV absorption curve of a PS containing styrene monomer Reproduced from Crompton and co-workers, British Plastics [76]

Table 2.12 Comparison of direct UV and distillation/UV methods for the determination of styrene monomer Method

Solvent

Styrene monomer (% w/w) Polystyrene sample No. 1

No. 2

Direct UV

Chloroform Carbon tetrachloride Ethyl acetate

Determination of Additives in Polymers and Rubber

Additives in Polymers

Analysis of rubber and rubber-like polymers

Additives in Polymers.. Industrial Analysis and Applications

NMR Spectra of Polymers and Polymer Additives

NMR Spectra of Polymers and Polymer Additives

NMR Spectra of Polymers and Polymer Additives

Rubber Analysis ї Polymers, Compounds and Products

Rubber and Rubber Balloons

Rubber

Rubber

Rubber

Rubber

Rubber

Rubber

A Concise Introduction to Additives for Thermoplastic Polymers (Wiley-Scrivener)

FUELS AND FUEL-ADDITIVES

Handbook of Food Additives

Atlas of Plastics Additives

Blends of Natural Rubber

Physical Testing of Rubber

The Romance Of Rubber

Modeling and Simulation in Polymers

Physics of photorefraction in polymers

Polymers in Building and Construction

Physics of Photorefraction in Polymers

Handbook of Rubber Bonding

Mixing of Rubber

Handbook of Polymers in Electronics

Plastics, Rubber and Health

Determination of Additives in Polymers and Rubber

Additives in Polymers

Analysis of rubber and rubber-like polymers

Additives in Polymers.. Industrial Analysis and Applications

NMR Spectra of Polymers and Polymer Additives

NMR Spectra of Polymers and Polymer Additives

NMR Spectra of Polymers and Polymer Additives

Rubber Analysis ї Polymers, Compounds and Products

Rubber and Rubber Balloons

Rubber

Rubber

Rubber

Rubber

Rubber

Rubber

A Concise Introduction to Additives for Thermoplastic Polymers (Wiley-Scrivener)

FUELS AND FUEL-ADDITIVES

Handbook of Food Additives

Atlas of Plastics Additives

Blends of Natural Rubber

Physical Testing of Rubber

The Romance Of Rubber

Modeling and Simulation in Polymers

Physics of photorefraction in polymers

Polymers in Building and Construction

Physics of Photorefraction in Polymers

Handbook of Rubber Bonding

Mixing of Rubber

Handbook of Polymers in Electronics

Plastics, Rubber and Health

Recommend Documents